Ch14_Computer_Vision/Single_Shot_Multibox_Detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Single Shot Multibox Detection (SSD)\n",
    "\n",
    "In the previous few sections, we have introduced bounding boxes, anchor boxes,\n",
    "multiscale object detection, and data sets. Now, we will use this background\n",
    "knowledge to construct an object detection model: single shot multibox detection\n",
    "(SSD) :cite:`Liu.Anguelov.Erhan.ea.2016`. This quick and easy model is already\n",
    "widely used. Some of the design concepts and implementation details of this\n",
    "model are also applicable to other object detection models.\n",
    "\n",
    "\n",
    "## Model\n",
    "\n",
    ":numref:`fig_ssd` shows the design of an SSD model. The model's main components\n",
    "are a base network block and several multiscale feature blocks connected in a\n",
    "series. Here, the base network block is used to extract features of original\n",
    "images, and it generally takes the form of a deep convolutional neural\n",
    "network. The paper on SSDs chooses to place a truncated VGG before the\n",
    "classification layer :cite:`Liu.Anguelov.Erhan.ea.2016`, but this is now\n",
    "commonly replaced by ResNet. We can design the base network so that it outputs\n",
    "larger heights and widths. In this way, more anchor boxes are generated based on\n",
    "this feature map, allowing us to detect smaller objects. Next, each multiscale\n",
    "feature block reduces the height and width of the feature map provided by the\n",
    "previous layer (for example, it may reduce the sizes by half). The blocks then\n",
    "use each element in the feature map to expand the receptive field on the input\n",
    "image. In this way, the closer a multiscale feature block is to the top of\n",
    ":numref:`fig_ssd` the smaller its output feature map, and the fewer the anchor\n",
    "boxes that are generated based on the feature map. In addition, the closer a\n",
    "feature block is to the top, the larger the receptive field of each element in\n",
    "the feature map and the better suited it is to detect larger objects. As the SSD\n",
    "generates different numbers of anchor boxes of different sizes based on the base\n",
    "network block and each multiscale feature block and then predicts the categories\n",
    "and offsets (i.e., predicted bounding boxes) of the anchor boxes in order to\n",
    "detect objects of different sizes, SSD is a multiscale object detection model.\n",
    "\n",
    "![The SSD is composed of a base network block and several multiscale feature blocks connected in a series. ](../img/ssd.svg)\n",
    "\n",
    ":label:`fig_ssd`\n",
    "\n",
    "\n",
    "\n",
    "Next, we will describe the implementation of the modules in :numref:`fig_ssd`. First, we need to discuss the implementation of category prediction and bounding box prediction.\n",
    "\n",
    "### Category Prediction Layer\n",
    "\n",
    "Set the number of object categories to $q$. In this case, the number of anchor\n",
    "box categories is $q+1$, with 0 indicating an anchor box that only contains\n",
    "background. For a certain scale, set the height and width of the feature map to\n",
    "$h$ and $w$, respectively. If we use each element as the center to generate $a$\n",
    "anchor boxes, we need to classify a total of $hwa$ anchor boxes. If we use a\n",
    "fully connected layer (FCN) for the output, this will likely result in an\n",
    "excessive number of model parameters. Recall how we used convolutional layer\n",
    "channels to output category predictions in :numref:`chapter_nin`. SSD uses the\n",
    "same method to reduce the model complexity.\n",
    "\n",
    "Specifically, the category prediction layer uses a convolutional layer that\n",
    "maintains the input height and width. Thus, the output and input have a\n",
    "one-to-one correspondence to the spatial coordinates along the width and height\n",
    "of the feature map. Assuming that the output and input have the same spatial\n",
    "coordinates $(x,y)$, the channel for the coordinates $(x,y)$ on the output\n",
    "feature map contains the category predictions for all anchor boxes generated\n",
    "using the input feature map coordinates $(x,y)$ as the center. Therefore, there\n",
    "are $a(q+1)$ output channels, with the output channels indexed as $i(q+1) + j$\n",
    "($0 \\leq j \\leq q$) representing the predictions of the category index $j$ for\n",
    "the anchor box index $i$.\n",
    "\n",
    "Now, we will define a category prediction layer of this type. After we specify\n",
    "the parameters $a$ and $q$, it uses a $3\\times3$ convolutional layer with a\n",
    "padding of 1. The heights and widths of the input and output of this\n",
    "convolutional layer remain unchanged."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import sys\n",
    "sys.path.insert(0, '..')\n",
    "import d2l\n",
    "# from d2l.ssd_utils import *\n",
    "import numpy as np\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "import torchvision.transforms as transforms\n",
    "import json\n",
    "import time\n",
    "from tqdm import tqdm\n",
    "from PIL import Image\n",
    "def cls_predictor(input_channels, num_anchors, num_classes):\n",
    "    return nn.Conv2d(in_channels=input_channels, out_channels=num_anchors * (num_classes + 1), kernel_size=3,\n",
    "                     padding=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bounding Box Prediction Layer\n",
    "\n",
    "The design of the bounding box prediction layer is similar to that of the category prediction layer. The only difference is that, here, we need to predict 4 offsets for each anchor box, rather than $q+1$ categories."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bbox_predictor(input_channels, num_anchors):\n",
    "    return nn.Conv2d(in_channels=input_channels, out_channels=num_anchors * 4, kernel_size=3, padding=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Concatenating Predictions for Multiple Scales\n",
    "\n",
    "As we mentioned, SSD uses feature maps based on multiple scales to generate anchor boxes and predict their categories and offsets. Because the shapes and number of anchor boxes centered on the same element differ for the feature maps of different scales, the prediction outputs at different scales may have different shapes.\n",
    "\n",
    "In the following example, we use the same batch of data to construct feature maps of two different scales, `Y1` and `Y2`. Here, `Y2` has half the height and half the width of `Y1`. Using category prediction as an example, we assume that each element in the `Y1` and `Y2` feature maps generates five (Y1) or three (Y2) anchor boxes. When there are 10 object categories, the number of category prediction output channels is either $5\\times(10+1)=55$ or $3\\times(10+1)=33$. The format of the prediction output is (batch size, number of channels, height, width). As you can see, except for the batch size, the sizes of the other dimensions are different. Therefore, we must transform them into a consistent format and concatenate the predictions of the multiple scales to facilitate subsequent computation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([2, 55, 20, 20]), torch.Size([2, 33, 10, 10]))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def forward(x, block):\n",
    "    return block(x)\n",
    "Y1 = forward(torch.zeros((2, 8, 20, 20)), cls_predictor(8, 5, 10))\n",
    "Y2 = forward(torch.zeros((2, 16, 10, 10)), cls_predictor(16, 3, 10))\n",
    "(Y1.shape, Y2.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The channel dimension contains the predictions for all anchor boxes with the same center. We first move the channel dimension to the final dimension. Because the batch size is the same for all scales, we can convert the prediction results to binary format (batch size, height $\\times$ width $\\times$ number of channels) to facilitate subsequent concatenation on the 1st dimension."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def flatten_pred(pred):\n",
    "    return pred.permute(0, 2, 3, 1).reshape(pred.size(0),-1)\n",
    "\n",
    "def concat_preds(preds):\n",
    "    return torch.cat(tuple([flatten_pred(p) for p in preds]), dim=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus, regardless of the different shapes of `Y1` and `Y2`, we can still concatenate the prediction results for the two different scales of the same batch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([2, 25300])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "concat_preds([Y1, Y2]).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Height and Width Downsample Block\n",
    "\n",
    "For multiscale object detection, we define the following `down_sample_blk` block, which reduces the height and width by 50%. This block consists of two $3\\times3$ convolutional layers with a padding of 1 and a $2\\times2$ maximum pooling layer with a stride of 2 connected in a series. As we know, $3\\times3$ convolutional layers with a padding of 1 do not change the shape of feature maps. However, the subsequent pooling layer directly reduces the size of the feature map by half. Because $1\\times 2+(3-1)+(3-1)=6$, each element in the output feature map has a receptive field on the input feature map of the shape $6\\times6$. As you can see, the height and width downsample block enlarges the receptive field of each element in the output feature map."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def down_sample_blk(input_channels, num_channels):\n",
    "    blk = []\n",
    "    for _ in range(2):\n",
    "        blk.append(nn.Conv2d(in_channels=input_channels, out_channels=num_channels, kernel_size=3, padding=1))\n",
    "        blk.append(nn.BatchNorm2d(num_features=num_channels))\n",
    "        blk.append(nn.ReLU())\n",
    "        input_channels=num_channels\n",
    "    blk.append(nn.MaxPool2d(kernel_size=2, stride=2))\n",
    "    blk = nn.Sequential(*blk)\n",
    "    return blk"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By testing forward computation in the height and width downsample block, we can see that it changes the number of input channels and halves the height and width."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([2, 10, 10, 10])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forward(torch.zeros((2, 3, 20, 20)), down_sample_blk(3, 10)).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Base Network Block\n",
    "\n",
    "The base network block is used to extract features from original images. To simplify the computation, we will construct a small base network. This network consists of three height and width downsample blocks connected in a series, so it doubles the number of channels at each step. When we input an original image with the shape $256\\times256$, the base network block outputs a feature map with the shape $32 \\times 32$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([2, 64, 32, 32])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def base_net():\n",
    "    blk = []\n",
    "    num_filters = [3, 16, 32, 64]\n",
    "    for i in range(len(num_filters) - 1):\n",
    "        blk.append(down_sample_blk(num_filters[i], num_filters[i+1]))\n",
    "    blk = nn.Sequential(*blk)\n",
    "    return blk\n",
    "\n",
    "forward(torch.zeros((2, 3, 256, 256)), base_net()).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The Complete Model\n",
    "\n",
    "The SSD model contains a total of five modules. Each module outputs a feature\n",
    "map used to generate anchor boxes and predict the categories and offsets of\n",
    "these anchor boxes. The first module is the base network block, modules two to\n",
    "four are height and width downsample blocks, and the fifth module is a global\n",
    "maximum pooling layer that reduces the height and width to 1. Therefore, modules\n",
    "two to five are all multiscale feature blocks shown in :numref:`fig_ssd`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_blk(i):\n",
    "    if i == 0:\n",
    "        blk = base_net()\n",
    "    elif i == 1:\n",
    "        blk = down_sample_blk(64, 128)\n",
    "    elif i == 4:\n",
    "        blk = nn.AdaptiveMaxPool2d((1,1))\n",
    "    else:\n",
    "        blk = down_sample_blk(128, 128)\n",
    "    \n",
    "    return blk"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we will define the forward computation process for each module. In contrast to the previously-described convolutional neural networks, this module not only returns feature map `Y` output by convolutional computation, but also the anchor boxes of the current scale generated from `Y` and their predicted categories and offsets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import itertools\n",
    "import math\n",
    "def create_anchors(feature_map_sizes, steps, sizes):\n",
    "    \"\"\"Compute default box sizes with scale and aspect transform.\"\"\"\n",
    "    scale = 256.\n",
    "    steps = [s / scale for s in steps]\n",
    "    sizes = [s / scale for s in sizes]\n",
    "    \n",
    "    aspect_ratios = ((2,),)\n",
    "\n",
    "\n",
    "    num_layers = len(feature_map_sizes)\n",
    "\n",
    "    boxes = []\n",
    "    for i in range(num_layers):\n",
    "        fmsize = feature_map_sizes[i]\n",
    "        for h, w in itertools.product(range(fmsize), repeat=2):\n",
    "            cx = (w + 0.5)*steps[i]\n",
    "            cy = (h + 0.5)*steps[i]\n",
    "            s = sizes[i]\n",
    "            boxes.append((cx, cy, s, s))\n",
    "\n",
    "            s = sizes[i+1]\n",
    "            boxes.append((cx, cy, s, s))\n",
    "\n",
    "            s = sizes[i]\n",
    "            for ar in aspect_ratios[i]:\n",
    "                \n",
    "#                 boxes.append((cx - (s * math.sqrt(ar))/2, cy - (s / math.sqrt(ar))/2, cx + (s * math.sqrt(ar))/2, cy + (s / math.sqrt(ar))/2))\n",
    "#                 boxes.append((cx - (s / math.sqrt(ar))/2, cy - (s * math.sqrt(ar))/2, cx + (s / math.sqrt(ar))/2, cy + (s * math.sqrt(ar))/2))\n",
    "\n",
    "                boxes.append((cx, cy, (s * math.sqrt(ar)), (s / math.sqrt(ar))))\n",
    "                boxes.append((cx, cy, (s / math.sqrt(ar)), (s * math.sqrt(ar))))\n",
    "\n",
    "    return torch.Tensor(boxes) # [8632, 4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def blk_forward(X, blk, size, ratio, cls_predictor, bbox_predictor):\n",
    "    Y = blk(X)\n",
    "    anchors = create_anchors((Y.size(2),), (256/Y.size(2),), size)\n",
    "    cls_preds = cls_predictor(Y)\n",
    "    bbox_preds = bbox_predictor(Y)\n",
    "    return (Y, anchors, cls_preds, bbox_preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we mentioned, the closer a multiscale feature block is to the top in :numref:`fig_ssd`, the larger the objects it detects and the larger the anchor boxes it must generate. Here, we first divide the interval from 0.2 to 1.05 into five equal parts to determine the sizes of smaller anchor boxes at different scales: 0.2, 0.37, 0.54, etc. Then, according to $\\sqrt{0.2 \\times 0.37} = 0.272$, $\\sqrt{0.37 \\times 0.54} = 0.447$, and similar formulas, we determine the sizes of larger anchor boxes at the different scales."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "sizes = [[0.2*256, 0.272*256], [0.37*256, 0.447*256], [0.54*256, 0.619*256],\n",
    "         [0.71*256, 0.79*256], [0.88*256, 0.961*256]]\n",
    "ratios = [[1, 2, 0.5]] * 5\n",
    "num_anchors = len(sizes[0]) + len(ratios[0]) - 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we can define the complete model, `TinySSD`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "class TinySSD(nn.Module):\n",
    "    def __init__(self, input_channels, num_classes):\n",
    "        super(TinySSD, self).__init__()\n",
    "        \n",
    "        input_channels_cls = 128\n",
    "        input_channels_bbox = 128\n",
    "        self.num_classes = num_classes\n",
    "        \n",
    "        self.blk = []\n",
    "        self.cls = []\n",
    "        self.bbox = []\n",
    "        \n",
    "        self.blk_0 = get_blk(0)\n",
    "        self.blk_1 = get_blk(1)\n",
    "        self.blk_2 = get_blk(2)\n",
    "        self.blk_3 = get_blk(3)\n",
    "        self.blk_4 = get_blk(4)\n",
    "        \n",
    "        self.cls_0 = cls_predictor(64, num_anchors, num_classes)\n",
    "        self.cls_1 = cls_predictor(input_channels_cls, num_anchors, num_classes)\n",
    "        self.cls_2 = cls_predictor(input_channels_cls, num_anchors, num_classes)\n",
    "        self.cls_3 = cls_predictor(input_channels_cls, num_anchors, num_classes)\n",
    "        self.cls_4 = cls_predictor(input_channels_cls, num_anchors, num_classes)\n",
    "        \n",
    "        self.bbox_0 = bbox_predictor(64, num_anchors)\n",
    "        self.bbox_1 = bbox_predictor(input_channels_bbox, num_anchors)\n",
    "        self.bbox_2 = bbox_predictor(input_channels_bbox, num_anchors)\n",
    "        self.bbox_3 = bbox_predictor(input_channels_bbox, num_anchors)\n",
    "        self.bbox_4 = bbox_predictor(input_channels_bbox, num_anchors)\n",
    "    \n",
    "    def forward(self, X):\n",
    "        anchors, cls_preds, bbox_preds = [None] * 5, [None] * 5, [None] * 5\n",
    "        \n",
    "        X, anchors[0], cls_preds[0], bbox_preds[0] = blk_forward(X, self.blk_0, sizes[0], ratios[0],\n",
    "                                                                    self.cls_0, self.bbox_0)\n",
    "        \n",
    "        X, anchors[1], cls_preds[1], bbox_preds[1] = blk_forward(X, self.blk_1, sizes[1], ratios[1],\n",
    "                                                                    self.cls_1, self.bbox_1)\n",
    "            \n",
    "        X, anchors[2], cls_preds[2], bbox_preds[2] = blk_forward(X, self.blk_2, sizes[2], ratios[2],\n",
    "                                                                    self.cls_2, self.bbox_2)    \n",
    "        \n",
    "        X, anchors[3], cls_preds[3], bbox_preds[3] = blk_forward(X, self.blk_3, sizes[3], ratios[3],\n",
    "                                                                    self.cls_3, self.bbox_3)    \n",
    "        \n",
    "        X, anchors[4], cls_preds[4], bbox_preds[4] = blk_forward(X, self.blk_4, sizes[4], ratios[4],\n",
    "                                                                    self.cls_4, self.bbox_4)    \n",
    "\n",
    "        return (torch.cat(anchors, dim=0), concat_preds(cls_preds).reshape((-1, 5444, self.num_classes + 1)), \n",
    "                concat_preds(bbox_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "anchors, cls_preds, bbox_preds = [None] * 5, [None] * 5, [None] * 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now create an SSD model instance and use it to perform forward computation on image mini-batch `X`, which has a height and width of 256 pixels. As we verified previously, the first module outputs a feature map with the shape $32 \\times 32$. Because modules two to four are height and width downsample blocks, module five is a global pooling layer, and each element in the feature map is used as the center for 4 anchor boxes, a total of $(32^2 + 16^2 + 8^2 + 4^2 + 1)\\times 4 = 5444$ anchor boxes are generated for each image at the five scales."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "output anchors: torch.Size([5444, 4])\n",
      "output class preds: torch.Size([32, 5444, 2])\n",
      "output bbox preds: torch.Size([32, 21776])\n"
     ]
    }
   ],
   "source": [
    "def init_weights(m):\n",
    "    if type(m) == nn.Linear or type(m) == nn.Conv2d:\n",
    "        torch.nn.init.xavier_uniform_(m.weight)\n",
    "        \n",
    "net = TinySSD(3, num_classes=1)\n",
    "net.apply(init_weights)\n",
    "\n",
    "X = torch.zeros((32, 3, 256, 256))\n",
    "anchors, cls_preds, bbox_preds = net(X)\n",
    "\n",
    "print('output anchors:', anchors.shape)\n",
    "print('output class preds:', cls_preds.shape)\n",
    "print('output bbox preds:', bbox_preds.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training\n",
    "\n",
    "Now, we will explain, step by step, how to train the SSD model for object detection.\n",
    "\n",
    "### Data Reading and Initialization\n",
    "\n",
    "We read the Pikachu data set we created in the previous section."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this part you need mxnet library to download and convert the mxnet's .rec file to png images and they will be saved in the 'data' directory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "d2l.download_and_preprocess_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "data_dir = '../data/pikachu/'\n",
    "train_dataset = d2l.PIKACHU(data_dir, 'train')\n",
    "val_dataset = d2l.PIKACHU(data_dir, 'val')\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(train_dataset,\n",
    "                                           batch_size=batch_size, shuffle=True,\n",
    "                                           num_workers=4)\n",
    "\n",
    "val_loader = torch.utils.data.DataLoader(val_dataset,\n",
    "                                         batch_size=batch_size, shuffle=False,\n",
    "                                         num_workers=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is 1 category in the Pikachu data set. After defining the module, we need to initialize the model parameters and define the optimization algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import os\n",
    "# os.environ[\"CUDA_VISIBLE_DEVICES\"] = '7' # You can set gpu no. if you have multiple gpus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda:0\n"
     ]
    }
   ],
   "source": [
    "device = torch.device('cuda:0') if torch.cuda.is_available() else torch.device('cpu')\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "net = TinySSD(3, num_classes=1)\n",
    "net.apply(init_weights)\n",
    "net = net.to(device)\n",
    "\n",
    "learning_rate = 1e-3\n",
    "weight_decay = 5e-4\n",
    "optimizer = optim.SGD(net.parameters(), lr = learning_rate, weight_decay=weight_decay)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define Loss and Evaluation Functions\n",
    "\n",
    "Object detection is subject to two types of losses. The first is anchor box category loss. For this, we can simply reuse the cross-entropy loss function we used in image classification. The second loss is positive anchor box offset loss. Offset prediction is a normalization problem. However, here, we do not use the squared loss introduced previously. Rather, we use the $L_1$ norm loss, which is the absolute value of the difference between the predicted value and the ground-truth value.Then we remove the negative anchor boxes and padding anchor boxes from the loss calculation. Finally, we add the anchor box category and offset losses to find the final loss function for the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use the accuracy rate to evaluate the classification results. As we use the $L_1$ norm loss, we will use the average absolute error to evaluate the bounding box prediction results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_cat = dict()\n",
    "id_cat[0] = 'pikachu'\n",
    "\n",
    "class FocalLoss(nn.Module):\n",
    "    def __init__(self, alpha=0.25, gamma=2, device=\"cuda:0\", eps=1e-10):\n",
    "        super().__init__()\n",
    "        self.alpha = alpha\n",
    "        self.gamma = gamma\n",
    "        self.device = device\n",
    "        self.eps = eps\n",
    "\n",
    "    def forward(self, input, target):\n",
    "        p = torch.sigmoid(input)\n",
    "        pt = p * target.float() + (1.0 - p) * (1 - target).float()\n",
    "        alpha_t = (1.0 - self.alpha) * target.float() + self.alpha * (1 - target).float()\n",
    "        loss = - 1.0 * torch.pow((1 - pt), self.gamma) * torch.log(pt + self.eps)\n",
    "        return loss.sum()\n",
    "    \n",
    "class SSDLoss(nn.Module):\n",
    "    def __init__(self, loc_factor, jaccard_overlap, device = \"cuda:0\", **kwargs):\n",
    "        super().__init__()\n",
    "        self.fl = FocalLoss(**kwargs)\n",
    "        self.loc_factor = loc_factor\n",
    "        self.jaccard_overlap = jaccard_overlap\n",
    "        \n",
    "\n",
    "        \n",
    "        self.device = device\n",
    "\n",
    "    def one_hot_encoding(labels, num_classes):\n",
    "        return torch.eye(num_classes)[labels]\n",
    "\n",
    "    def loc_transformation(x, anchors, overlap_indicies):\n",
    "        # Doing location transformations according to SSD paper\n",
    "        return torch.cat([(x[:, 0:1] - anchors[overlap_indicies, 0:1]) / anchors[overlap_indicies, 2:3],\n",
    "                          (x[:, 1:2] - anchors[overlap_indicies, 1:2]) / anchors[overlap_indicies, 3:4],\n",
    "                          torch.log((x[:, 2:3] / anchors[overlap_indicies, 2:3])),\n",
    "                          torch.log((x[:, 3:4] / anchors[overlap_indicies, 3:4]))\n",
    "                         ], dim=1)\n",
    "\n",
    "    def forward(self, class_hat, bb_hat, class_true, bb_true, anchors):        \n",
    "        loc_loss = 0.0\n",
    "        class_loss = 0.0\n",
    "    \n",
    "        \n",
    "        for i in range(len(class_true)):  # Batch level\n",
    "\n",
    "            class_hat_i = class_hat[i, :, :]\n",
    "\n",
    "            bb_true_i = bb_true[i].float()\n",
    "            \n",
    "            class_true_i = class_true[i]            \n",
    "            \n",
    "            class_target = torch.zeros(class_hat_i.shape[0]).long().to(self.device)\n",
    "            \n",
    "            overlap_list = d2l.find_overlap(bb_true_i.squeeze(0), anchors, self.jaccard_overlap)\n",
    "            \n",
    "            temp_loc_loss = 0.0\n",
    "            for j in range(len(overlap_list)):  # BB level\n",
    "                overlap = overlap_list[j]\n",
    "                class_target[overlap] = class_true_i[0, j].long()\n",
    "\n",
    "                input_ = bb_hat[i, overlap, :]\n",
    "                target_ = SSDLoss.loc_transformation(bb_true_i[0, j, :].expand((len(overlap), 4)), anchors, overlap)\n",
    "\n",
    "                temp_loc_loss += F.smooth_l1_loss(input=input_, target=target_, reduction=\"sum\") / len(overlap)\n",
    "            loc_loss += temp_loc_loss / class_true_i.shape[1]\n",
    "\n",
    "            class_target = SSDLoss.one_hot_encoding(class_target, len(id_cat) + 1).float().to(self.device)\n",
    "            class_loss += self.fl(class_hat_i, class_target) / class_true_i.shape[1]\n",
    "\n",
    "        loc_loss = loc_loss / len(class_true)\n",
    "        class_loss = class_loss / len(class_true)\n",
    "        loss = class_loss + loc_loss * self.loc_factor\n",
    "\n",
    "        return loss, loc_loss, class_loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train the Model\n",
    "\n",
    "During model training, we must generate multiscale anchor boxes (`anchors`) in the model's forward computation process and predict the category (`cls_preds`) and offset (`bbox_preds`) for each anchor box.  Finally, we calculate the loss function using the predicted and labeled category and offset values. To simplify the code, we do not evaluate the training data set here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss = SSDLoss(loc_factor=5.0, jaccard_overlap=0.5, device=\"cuda:0\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (https://matplotlib.org/) -->\n",
       "<svg height=\"180.65625pt\" version=\"1.1\" viewBox=\"0 0 242.15 180.65625\" width=\"242.15pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 180.65625 \n",
       "L 242.15 180.65625 \n",
       "L 242.15 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 33.2875 143.1 \n",
       "L 228.5875 143.1 \n",
       "L 228.5875 7.2 \n",
       "L 33.2875 7.2 \n",
       "z\n",
       "\" style=\"fill:#ffffff;\"/>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_1\">\n",
       "    <g id=\"xtick_1\">\n",
       "     <g id=\"line2d_1\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 143.1 \n",
       "L 33.2875 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_2\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L 0 3.5 \n",
       "\" id=\"m6b5453b3ff\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_1\">\n",
       "      <!-- 0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 31.78125 66.40625 \n",
       "Q 24.171875 66.40625 20.328125 58.90625 \n",
       "Q 16.5 51.421875 16.5 36.375 \n",
       "Q 16.5 21.390625 20.328125 13.890625 \n",
       "Q 24.171875 6.390625 31.78125 6.390625 \n",
       "Q 39.453125 6.390625 43.28125 13.890625 \n",
       "Q 47.125 21.390625 47.125 36.375 \n",
       "Q 47.125 51.421875 43.28125 58.90625 \n",
       "Q 39.453125 66.40625 31.78125 66.40625 \n",
       "z\n",
       "M 31.78125 74.21875 \n",
       "Q 44.046875 74.21875 50.515625 64.515625 \n",
       "Q 56.984375 54.828125 56.984375 36.375 \n",
       "Q 56.984375 17.96875 50.515625 8.265625 \n",
       "Q 44.046875 -1.421875 31.78125 -1.421875 \n",
       "Q 19.53125 -1.421875 13.0625 8.265625 \n",
       "Q 6.59375 17.96875 6.59375 36.375 \n",
       "Q 6.59375 54.828125 13.0625 64.515625 \n",
       "Q 19.53125 74.21875 31.78125 74.21875 \n",
       "z\n",
       "\" id=\"DejaVuSans-48\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(30.10625 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_2\">\n",
       "     <g id=\"line2d_3\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 72.3475 143.1 \n",
       "L 72.3475 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_4\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"72.3475\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_2\">\n",
       "      <!-- 5 -->\n",
       "      <defs>\n",
       "       <path d=\"M 10.796875 72.90625 \n",
       "L 49.515625 72.90625 \n",
       "L 49.515625 64.59375 \n",
       "L 19.828125 64.59375 \n",
       "L 19.828125 46.734375 \n",
       "Q 21.96875 47.46875 24.109375 47.828125 \n",
       "Q 26.265625 48.1875 28.421875 48.1875 \n",
       "Q 40.625 48.1875 47.75 41.5 \n",
       "Q 54.890625 34.8125 54.890625 23.390625 \n",
       "Q 54.890625 11.625 47.5625 5.09375 \n",
       "Q 40.234375 -1.421875 26.90625 -1.421875 \n",
       "Q 22.3125 -1.421875 17.546875 -0.640625 \n",
       "Q 12.796875 0.140625 7.71875 1.703125 \n",
       "L 7.71875 11.625 \n",
       "Q 12.109375 9.234375 16.796875 8.0625 \n",
       "Q 21.484375 6.890625 26.703125 6.890625 \n",
       "Q 35.15625 6.890625 40.078125 11.328125 \n",
       "Q 45.015625 15.765625 45.015625 23.390625 \n",
       "Q 45.015625 31 40.078125 35.4375 \n",
       "Q 35.15625 39.890625 26.703125 39.890625 \n",
       "Q 22.75 39.890625 18.8125 39.015625 \n",
       "Q 14.890625 38.140625 10.796875 36.28125 \n",
       "z\n",
       "\" id=\"DejaVuSans-53\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(69.16625 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-53\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_3\">\n",
       "     <g id=\"line2d_5\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 111.4075 143.1 \n",
       "L 111.4075 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_6\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"111.4075\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_3\">\n",
       "      <!-- 10 -->\n",
       "      <defs>\n",
       "       <path d=\"M 12.40625 8.296875 \n",
       "L 28.515625 8.296875 \n",
       "L 28.515625 63.921875 \n",
       "L 10.984375 60.40625 \n",
       "L 10.984375 69.390625 \n",
       "L 28.421875 72.90625 \n",
       "L 38.28125 72.90625 \n",
       "L 38.28125 8.296875 \n",
       "L 54.390625 8.296875 \n",
       "L 54.390625 0 \n",
       "L 12.40625 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-49\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(105.045 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_4\">\n",
       "     <g id=\"line2d_7\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 150.4675 143.1 \n",
       "L 150.4675 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_8\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"150.4675\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_4\">\n",
       "      <!-- 15 -->\n",
       "      <g transform=\"translate(144.105 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_5\">\n",
       "     <g id=\"line2d_9\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 189.5275 143.1 \n",
       "L 189.5275 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_10\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"189.5275\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_5\">\n",
       "      <!-- 20 -->\n",
       "      <defs>\n",
       "       <path d=\"M 19.1875 8.296875 \n",
       "L 53.609375 8.296875 \n",
       "L 53.609375 0 \n",
       "L 7.328125 0 \n",
       "L 7.328125 8.296875 \n",
       "Q 12.9375 14.109375 22.625 23.890625 \n",
       "Q 32.328125 33.6875 34.8125 36.53125 \n",
       "Q 39.546875 41.84375 41.421875 45.53125 \n",
       "Q 43.3125 49.21875 43.3125 52.78125 \n",
       "Q 43.3125 58.59375 39.234375 62.25 \n",
       "Q 35.15625 65.921875 28.609375 65.921875 \n",
       "Q 23.96875 65.921875 18.8125 64.3125 \n",
       "Q 13.671875 62.703125 7.8125 59.421875 \n",
       "L 7.8125 69.390625 \n",
       "Q 13.765625 71.78125 18.9375 73 \n",
       "Q 24.125 74.21875 28.421875 74.21875 \n",
       "Q 39.75 74.21875 46.484375 68.546875 \n",
       "Q 53.21875 62.890625 53.21875 53.421875 \n",
       "Q 53.21875 48.921875 51.53125 44.890625 \n",
       "Q 49.859375 40.875 45.40625 35.40625 \n",
       "Q 44.1875 33.984375 37.640625 27.21875 \n",
       "Q 31.109375 20.453125 19.1875 8.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-50\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(183.165 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_6\">\n",
       "     <g id=\"line2d_11\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 228.5875 143.1 \n",
       "L 228.5875 7.2 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_12\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"228.5875\" xlink:href=\"#m6b5453b3ff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_6\">\n",
       "      <!-- 25 -->\n",
       "      <g transform=\"translate(222.225 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"text_7\">\n",
       "     <!-- epoch -->\n",
       "     <defs>\n",
       "      <path d=\"M 56.203125 29.59375 \n",
       "L 56.203125 25.203125 \n",
       "L 14.890625 25.203125 \n",
       "Q 15.484375 15.921875 20.484375 11.0625 \n",
       "Q 25.484375 6.203125 34.421875 6.203125 \n",
       "Q 39.59375 6.203125 44.453125 7.46875 \n",
       "Q 49.3125 8.734375 54.109375 11.28125 \n",
       "L 54.109375 2.78125 \n",
       "Q 49.265625 0.734375 44.1875 -0.34375 \n",
       "Q 39.109375 -1.421875 33.890625 -1.421875 \n",
       "Q 20.796875 -1.421875 13.15625 6.1875 \n",
       "Q 5.515625 13.8125 5.515625 26.8125 \n",
       "Q 5.515625 40.234375 12.765625 48.109375 \n",
       "Q 20.015625 56 32.328125 56 \n",
       "Q 43.359375 56 49.78125 48.890625 \n",
       "Q 56.203125 41.796875 56.203125 29.59375 \n",
       "z\n",
       "M 47.21875 32.234375 \n",
       "Q 47.125 39.59375 43.09375 43.984375 \n",
       "Q 39.0625 48.390625 32.421875 48.390625 \n",
       "Q 24.90625 48.390625 20.390625 44.140625 \n",
       "Q 15.875 39.890625 15.1875 32.171875 \n",
       "z\n",
       "\" id=\"DejaVuSans-101\"/>\n",
       "      <path d=\"M 18.109375 8.203125 \n",
       "L 18.109375 -20.796875 \n",
       "L 9.078125 -20.796875 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.390625 \n",
       "Q 20.953125 51.265625 25.265625 53.625 \n",
       "Q 29.59375 56 35.59375 56 \n",
       "Q 45.5625 56 51.78125 48.09375 \n",
       "Q 58.015625 40.1875 58.015625 27.296875 \n",
       "Q 58.015625 14.40625 51.78125 6.484375 \n",
       "Q 45.5625 -1.421875 35.59375 -1.421875 \n",
       "Q 29.59375 -1.421875 25.265625 0.953125 \n",
       "Q 20.953125 3.328125 18.109375 8.203125 \n",
       "z\n",
       "M 48.6875 27.296875 \n",
       "Q 48.6875 37.203125 44.609375 42.84375 \n",
       "Q 40.53125 48.484375 33.40625 48.484375 \n",
       "Q 26.265625 48.484375 22.1875 42.84375 \n",
       "Q 18.109375 37.203125 18.109375 27.296875 \n",
       "Q 18.109375 17.390625 22.1875 11.75 \n",
       "Q 26.265625 6.109375 33.40625 6.109375 \n",
       "Q 40.53125 6.109375 44.609375 11.75 \n",
       "Q 48.6875 17.390625 48.6875 27.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-112\"/>\n",
       "      <path d=\"M 30.609375 48.390625 \n",
       "Q 23.390625 48.390625 19.1875 42.75 \n",
       "Q 14.984375 37.109375 14.984375 27.296875 \n",
       "Q 14.984375 17.484375 19.15625 11.84375 \n",
       "Q 23.34375 6.203125 30.609375 6.203125 \n",
       "Q 37.796875 6.203125 41.984375 11.859375 \n",
       "Q 46.1875 17.53125 46.1875 27.296875 \n",
       "Q 46.1875 37.015625 41.984375 42.703125 \n",
       "Q 37.796875 48.390625 30.609375 48.390625 \n",
       "z\n",
       "M 30.609375 56 \n",
       "Q 42.328125 56 49.015625 48.375 \n",
       "Q 55.71875 40.765625 55.71875 27.296875 \n",
       "Q 55.71875 13.875 49.015625 6.21875 \n",
       "Q 42.328125 -1.421875 30.609375 -1.421875 \n",
       "Q 18.84375 -1.421875 12.171875 6.21875 \n",
       "Q 5.515625 13.875 5.515625 27.296875 \n",
       "Q 5.515625 40.765625 12.171875 48.375 \n",
       "Q 18.84375 56 30.609375 56 \n",
       "z\n",
       "\" id=\"DejaVuSans-111\"/>\n",
       "      <path d=\"M 48.78125 52.59375 \n",
       "L 48.78125 44.1875 \n",
       "Q 44.96875 46.296875 41.140625 47.34375 \n",
       "Q 37.3125 48.390625 33.40625 48.390625 \n",
       "Q 24.65625 48.390625 19.8125 42.84375 \n",
       "Q 14.984375 37.3125 14.984375 27.296875 \n",
       "Q 14.984375 17.28125 19.8125 11.734375 \n",
       "Q 24.65625 6.203125 33.40625 6.203125 \n",
       "Q 37.3125 6.203125 41.140625 7.25 \n",
       "Q 44.96875 8.296875 48.78125 10.40625 \n",
       "L 48.78125 2.09375 \n",
       "Q 45.015625 0.34375 40.984375 -0.53125 \n",
       "Q 36.96875 -1.421875 32.421875 -1.421875 \n",
       "Q 20.0625 -1.421875 12.78125 6.34375 \n",
       "Q 5.515625 14.109375 5.515625 27.296875 \n",
       "Q 5.515625 40.671875 12.859375 48.328125 \n",
       "Q 20.21875 56 33.015625 56 \n",
       "Q 37.15625 56 41.109375 55.140625 \n",
       "Q 45.0625 54.296875 48.78125 52.59375 \n",
       "z\n",
       "\" id=\"DejaVuSans-99\"/>\n",
       "      <path d=\"M 54.890625 33.015625 \n",
       "L 54.890625 0 \n",
       "L 45.90625 0 \n",
       "L 45.90625 32.71875 \n",
       "Q 45.90625 40.484375 42.875 44.328125 \n",
       "Q 39.84375 48.1875 33.796875 48.1875 \n",
       "Q 26.515625 48.1875 22.3125 43.546875 \n",
       "Q 18.109375 38.921875 18.109375 30.90625 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 75.984375 \n",
       "L 18.109375 75.984375 \n",
       "L 18.109375 46.1875 \n",
       "Q 21.34375 51.125 25.703125 53.5625 \n",
       "Q 30.078125 56 35.796875 56 \n",
       "Q 45.21875 56 50.046875 50.171875 \n",
       "Q 54.890625 44.34375 54.890625 33.015625 \n",
       "z\n",
       "\" id=\"DejaVuSans-104\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(115.709375 171.376563)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-101\"/>\n",
       "      <use x=\"61.523438\" xlink:href=\"#DejaVuSans-112\"/>\n",
       "      <use x=\"125\" xlink:href=\"#DejaVuSans-111\"/>\n",
       "      <use x=\"186.181641\" xlink:href=\"#DejaVuSans-99\"/>\n",
       "      <use x=\"241.162109\" xlink:href=\"#DejaVuSans-104\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_2\">\n",
       "    <g id=\"ytick_1\">\n",
       "     <g id=\"line2d_13\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 136.940346 \n",
       "L 228.5875 136.940346 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_14\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L -3.5 0 \n",
       "\" id=\"m553dbf7c33\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"136.940346\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_8\">\n",
       "      <!-- 0 -->\n",
       "      <g transform=\"translate(19.925 140.739564)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_2\">\n",
       "     <g id=\"line2d_15\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 111.968574 \n",
       "L 228.5875 111.968574 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_16\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"111.968574\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_9\">\n",
       "      <!-- 50 -->\n",
       "      <g transform=\"translate(13.5625 115.767793)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_3\">\n",
       "     <g id=\"line2d_17\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 86.996803 \n",
       "L 228.5875 86.996803 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_18\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"86.996803\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_10\">\n",
       "      <!-- 100 -->\n",
       "      <g transform=\"translate(7.2 90.796022)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_4\">\n",
       "     <g id=\"line2d_19\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 62.025031 \n",
       "L 228.5875 62.025031 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_20\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"62.025031\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_11\">\n",
       "      <!-- 150 -->\n",
       "      <g transform=\"translate(7.2 65.82425)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_5\">\n",
       "     <g id=\"line2d_21\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 37.05326 \n",
       "L 228.5875 37.05326 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_22\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"37.05326\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_12\">\n",
       "      <!-- 200 -->\n",
       "      <g transform=\"translate(7.2 40.852479)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_6\">\n",
       "     <g id=\"line2d_23\">\n",
       "      <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 12.081489 \n",
       "L 228.5875 12.081489 \n",
       "\" style=\"fill:none;stroke:#b0b0b0;stroke-linecap:square;stroke-width:0.8;\"/>\n",
       "     </g>\n",
       "     <g id=\"line2d_24\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#m553dbf7c33\" y=\"12.081489\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_13\">\n",
       "      <!-- 250 -->\n",
       "      <g transform=\"translate(7.2 15.880707)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"line2d_25\">\n",
       "    <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 14.813637 \n",
       "L 41.0995 119.336001 \n",
       "L 48.9115 125.270023 \n",
       "L 56.7235 128.784089 \n",
       "L 64.5355 130.466141 \n",
       "L 72.3475 130.10333 \n",
       "L 80.1595 130.570158 \n",
       "L 87.9715 130.142619 \n",
       "L 95.7835 131.638025 \n",
       "L 103.5955 132.310063 \n",
       "L 111.4075 132.210255 \n",
       "L 119.2195 132.480417 \n",
       "L 127.0315 132.683208 \n",
       "L 134.8435 133.140193 \n",
       "L 142.6555 132.261702 \n",
       "L 150.4675 131.558233 \n",
       "L 158.2795 132.876229 \n",
       "L 166.0915 133.459871 \n",
       "L 173.9035 133.445947 \n",
       "L 181.7155 133.841121 \n",
       "L 189.5275 133.687777 \n",
       "L 197.3395 133.560444 \n",
       "L 205.1515 133.876801 \n",
       "L 212.9635 134.124607 \n",
       "L 220.7755 134.197588 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_26\">\n",
       "    <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 136.461561 \n",
       "L 41.0995 136.839818 \n",
       "L 48.9115 136.847902 \n",
       "L 56.7235 136.853587 \n",
       "L 64.5355 136.858502 \n",
       "L 72.3475 136.874263 \n",
       "L 80.1595 136.884712 \n",
       "L 87.9715 136.896896 \n",
       "L 95.7835 136.892998 \n",
       "L 103.5955 136.892427 \n",
       "L 111.4075 136.895718 \n",
       "L 119.2195 136.907551 \n",
       "L 127.0315 136.908673 \n",
       "L 134.8435 136.912323 \n",
       "L 142.6555 136.91103 \n",
       "L 150.4675 136.914144 \n",
       "L 158.2795 136.913997 \n",
       "L 166.0915 136.915807 \n",
       "L 173.9035 136.917897 \n",
       "L 181.7155 136.918877 \n",
       "L 189.5275 136.918807 \n",
       "L 197.3395 136.917651 \n",
       "L 205.1515 136.920352 \n",
       "L 212.9635 136.922256 \n",
       "L 220.7755 136.922727 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_27\">\n",
       "    <path clip-path=\"url(#pd726c0490a)\" d=\"M 33.2875 13.377273 \n",
       "L 41.0995 119.03442 \n",
       "L 48.9115 124.992692 \n",
       "L 56.7235 128.523814 \n",
       "L 64.5355 130.220609 \n",
       "L 72.3475 129.905081 \n",
       "L 80.1595 130.403256 \n",
       "L 87.9715 130.01227 \n",
       "L 95.7835 131.495982 \n",
       "L 103.5955 132.166308 \n",
       "L 111.4075 132.076372 \n",
       "L 119.2195 132.382032 \n",
       "L 127.0315 132.588189 \n",
       "L 134.8435 133.056126 \n",
       "L 142.6555 132.173756 \n",
       "L 150.4675 131.479627 \n",
       "L 158.2795 132.797182 \n",
       "L 166.0915 133.386253 \n",
       "L 173.9035 133.378602 \n",
       "L 181.7155 133.776715 \n",
       "L 189.5275 133.623162 \n",
       "L 197.3395 133.492359 \n",
       "L 205.1515 133.81682 \n",
       "L 212.9635 134.070337 \n",
       "L 220.7755 134.144734 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 33.2875 143.1 \n",
       "L 33.2875 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_4\">\n",
       "    <path d=\"M 228.5875 143.1 \n",
       "L 228.5875 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_5\">\n",
       "    <path d=\"M 33.2875 143.1 \n",
       "L 228.5875 143.1 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_6\">\n",
       "    <path d=\"M 33.2875 7.2 \n",
       "L 228.5875 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"legend_1\">\n",
       "    <g id=\"patch_7\">\n",
       "     <path d=\"M 136.985938 59.5125 \n",
       "L 221.5875 59.5125 \n",
       "Q 223.5875 59.5125 223.5875 57.5125 \n",
       "L 223.5875 14.2 \n",
       "Q 223.5875 12.2 221.5875 12.2 \n",
       "L 136.985938 12.2 \n",
       "Q 134.985938 12.2 134.985938 14.2 \n",
       "L 134.985938 57.5125 \n",
       "Q 134.985938 59.5125 136.985938 59.5125 \n",
       "z\n",
       "\" style=\"fill:#ffffff;opacity:0.8;stroke:#cccccc;stroke-linejoin:miter;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_28\">\n",
       "     <path d=\"M 138.985938 20.298437 \n",
       "L 158.985938 20.298437 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_29\"/>\n",
       "    <g id=\"text_14\">\n",
       "     <!-- class error -->\n",
       "     <defs>\n",
       "      <path d=\"M 9.421875 75.984375 \n",
       "L 18.40625 75.984375 \n",
       "L 18.40625 0 \n",
       "L 9.421875 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-108\"/>\n",
       "      <path d=\"M 34.28125 27.484375 \n",
       "Q 23.390625 27.484375 19.1875 25 \n",
       "Q 14.984375 22.515625 14.984375 16.5 \n",
       "Q 14.984375 11.71875 18.140625 8.90625 \n",
       "Q 21.296875 6.109375 26.703125 6.109375 \n",
       "Q 34.1875 6.109375 38.703125 11.40625 \n",
       "Q 43.21875 16.703125 43.21875 25.484375 \n",
       "L 43.21875 27.484375 \n",
       "z\n",
       "M 52.203125 31.203125 \n",
       "L 52.203125 0 \n",
       "L 43.21875 0 \n",
       "L 43.21875 8.296875 \n",
       "Q 40.140625 3.328125 35.546875 0.953125 \n",
       "Q 30.953125 -1.421875 24.3125 -1.421875 \n",
       "Q 15.921875 -1.421875 10.953125 3.296875 \n",
       "Q 6 8.015625 6 15.921875 \n",
       "Q 6 25.140625 12.171875 29.828125 \n",
       "Q 18.359375 34.515625 30.609375 34.515625 \n",
       "L 43.21875 34.515625 \n",
       "L 43.21875 35.40625 \n",
       "Q 43.21875 41.609375 39.140625 45 \n",
       "Q 35.0625 48.390625 27.6875 48.390625 \n",
       "Q 23 48.390625 18.546875 47.265625 \n",
       "Q 14.109375 46.140625 10.015625 43.890625 \n",
       "L 10.015625 52.203125 \n",
       "Q 14.9375 54.109375 19.578125 55.046875 \n",
       "Q 24.21875 56 28.609375 56 \n",
       "Q 40.484375 56 46.34375 49.84375 \n",
       "Q 52.203125 43.703125 52.203125 31.203125 \n",
       "z\n",
       "\" id=\"DejaVuSans-97\"/>\n",
       "      <path d=\"M 44.28125 53.078125 \n",
       "L 44.28125 44.578125 \n",
       "Q 40.484375 46.53125 36.375 47.5 \n",
       "Q 32.28125 48.484375 27.875 48.484375 \n",
       "Q 21.1875 48.484375 17.84375 46.4375 \n",
       "Q 14.5 44.390625 14.5 40.28125 \n",
       "Q 14.5 37.15625 16.890625 35.375 \n",
       "Q 19.28125 33.59375 26.515625 31.984375 \n",
       "L 29.59375 31.296875 \n",
       "Q 39.15625 29.25 43.1875 25.515625 \n",
       "Q 47.21875 21.78125 47.21875 15.09375 \n",
       "Q 47.21875 7.46875 41.1875 3.015625 \n",
       "Q 35.15625 -1.421875 24.609375 -1.421875 \n",
       "Q 20.21875 -1.421875 15.453125 -0.5625 \n",
       "Q 10.6875 0.296875 5.421875 2 \n",
       "L 5.421875 11.28125 \n",
       "Q 10.40625 8.6875 15.234375 7.390625 \n",
       "Q 20.0625 6.109375 24.8125 6.109375 \n",
       "Q 31.15625 6.109375 34.5625 8.28125 \n",
       "Q 37.984375 10.453125 37.984375 14.40625 \n",
       "Q 37.984375 18.0625 35.515625 20.015625 \n",
       "Q 33.0625 21.96875 24.703125 23.78125 \n",
       "L 21.578125 24.515625 \n",
       "Q 13.234375 26.265625 9.515625 29.90625 \n",
       "Q 5.8125 33.546875 5.8125 39.890625 \n",
       "Q 5.8125 47.609375 11.28125 51.796875 \n",
       "Q 16.75 56 26.8125 56 \n",
       "Q 31.78125 56 36.171875 55.265625 \n",
       "Q 40.578125 54.546875 44.28125 53.078125 \n",
       "z\n",
       "\" id=\"DejaVuSans-115\"/>\n",
       "      <path id=\"DejaVuSans-32\"/>\n",
       "      <path d=\"M 41.109375 46.296875 \n",
       "Q 39.59375 47.171875 37.8125 47.578125 \n",
       "Q 36.03125 48 33.890625 48 \n",
       "Q 26.265625 48 22.1875 43.046875 \n",
       "Q 18.109375 38.09375 18.109375 28.8125 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.1875 \n",
       "Q 20.953125 51.171875 25.484375 53.578125 \n",
       "Q 30.03125 56 36.53125 56 \n",
       "Q 37.453125 56 38.578125 55.875 \n",
       "Q 39.703125 55.765625 41.0625 55.515625 \n",
       "z\n",
       "\" id=\"DejaVuSans-114\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(166.985938 23.798437)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-99\"/>\n",
       "      <use x=\"54.980469\" xlink:href=\"#DejaVuSans-108\"/>\n",
       "      <use x=\"82.763672\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"144.042969\" xlink:href=\"#DejaVuSans-115\"/>\n",
       "      <use x=\"196.142578\" xlink:href=\"#DejaVuSans-115\"/>\n",
       "      <use x=\"248.242188\" xlink:href=\"#DejaVuSans-32\"/>\n",
       "      <use x=\"280.029297\" xlink:href=\"#DejaVuSans-101\"/>\n",
       "      <use x=\"341.552734\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "      <use x=\"382.650391\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "      <use x=\"423.732422\" xlink:href=\"#DejaVuSans-111\"/>\n",
       "      <use x=\"484.914062\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_30\">\n",
       "     <path d=\"M 138.985938 34.976562 \n",
       "L 158.985938 34.976562 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_31\"/>\n",
       "    <g id=\"text_15\">\n",
       "     <!-- bbox mae -->\n",
       "     <defs>\n",
       "      <path d=\"M 48.6875 27.296875 \n",
       "Q 48.6875 37.203125 44.609375 42.84375 \n",
       "Q 40.53125 48.484375 33.40625 48.484375 \n",
       "Q 26.265625 48.484375 22.1875 42.84375 \n",
       "Q 18.109375 37.203125 18.109375 27.296875 \n",
       "Q 18.109375 17.390625 22.1875 11.75 \n",
       "Q 26.265625 6.109375 33.40625 6.109375 \n",
       "Q 40.53125 6.109375 44.609375 11.75 \n",
       "Q 48.6875 17.390625 48.6875 27.296875 \n",
       "z\n",
       "M 18.109375 46.390625 \n",
       "Q 20.953125 51.265625 25.265625 53.625 \n",
       "Q 29.59375 56 35.59375 56 \n",
       "Q 45.5625 56 51.78125 48.09375 \n",
       "Q 58.015625 40.1875 58.015625 27.296875 \n",
       "Q 58.015625 14.40625 51.78125 6.484375 \n",
       "Q 45.5625 -1.421875 35.59375 -1.421875 \n",
       "Q 29.59375 -1.421875 25.265625 0.953125 \n",
       "Q 20.953125 3.328125 18.109375 8.203125 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 75.984375 \n",
       "L 18.109375 75.984375 \n",
       "z\n",
       "\" id=\"DejaVuSans-98\"/>\n",
       "      <path d=\"M 54.890625 54.6875 \n",
       "L 35.109375 28.078125 \n",
       "L 55.90625 0 \n",
       "L 45.3125 0 \n",
       "L 29.390625 21.484375 \n",
       "L 13.484375 0 \n",
       "L 2.875 0 \n",
       "L 24.125 28.609375 \n",
       "L 4.6875 54.6875 \n",
       "L 15.28125 54.6875 \n",
       "L 29.78125 35.203125 \n",
       "L 44.28125 54.6875 \n",
       "z\n",
       "\" id=\"DejaVuSans-120\"/>\n",
       "      <path d=\"M 52 44.1875 \n",
       "Q 55.375 50.25 60.0625 53.125 \n",
       "Q 64.75 56 71.09375 56 \n",
       "Q 79.640625 56 84.28125 50.015625 \n",
       "Q 88.921875 44.046875 88.921875 33.015625 \n",
       "L 88.921875 0 \n",
       "L 79.890625 0 \n",
       "L 79.890625 32.71875 \n",
       "Q 79.890625 40.578125 77.09375 44.375 \n",
       "Q 74.3125 48.1875 68.609375 48.1875 \n",
       "Q 61.625 48.1875 57.5625 43.546875 \n",
       "Q 53.515625 38.921875 53.515625 30.90625 \n",
       "L 53.515625 0 \n",
       "L 44.484375 0 \n",
       "L 44.484375 32.71875 \n",
       "Q 44.484375 40.625 41.703125 44.40625 \n",
       "Q 38.921875 48.1875 33.109375 48.1875 \n",
       "Q 26.21875 48.1875 22.15625 43.53125 \n",
       "Q 18.109375 38.875 18.109375 30.90625 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.1875 \n",
       "Q 21.1875 51.21875 25.484375 53.609375 \n",
       "Q 29.78125 56 35.6875 56 \n",
       "Q 41.65625 56 45.828125 52.96875 \n",
       "Q 50 49.953125 52 44.1875 \n",
       "z\n",
       "\" id=\"DejaVuSans-109\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(166.985938 38.476562)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-98\"/>\n",
       "      <use x=\"63.476562\" xlink:href=\"#DejaVuSans-98\"/>\n",
       "      <use x=\"126.953125\" xlink:href=\"#DejaVuSans-111\"/>\n",
       "      <use x=\"188.087891\" xlink:href=\"#DejaVuSans-120\"/>\n",
       "      <use x=\"247.267578\" xlink:href=\"#DejaVuSans-32\"/>\n",
       "      <use x=\"279.054688\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"376.466797\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"437.746094\" xlink:href=\"#DejaVuSans-101\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_32\">\n",
       "     <path d=\"M 138.985938 49.654687 \n",
       "L 158.985938 49.654687 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_33\"/>\n",
       "    <g id=\"text_16\">\n",
       "     <!-- train_err -->\n",
       "     <defs>\n",
       "      <path d=\"M 18.3125 70.21875 \n",
       "L 18.3125 54.6875 \n",
       "L 36.8125 54.6875 \n",
       "L 36.8125 47.703125 \n",
       "L 18.3125 47.703125 \n",
       "L 18.3125 18.015625 \n",
       "Q 18.3125 11.328125 20.140625 9.421875 \n",
       "Q 21.96875 7.515625 27.59375 7.515625 \n",
       "L 36.8125 7.515625 \n",
       "L 36.8125 0 \n",
       "L 27.59375 0 \n",
       "Q 17.1875 0 13.234375 3.875 \n",
       "Q 9.28125 7.765625 9.28125 18.015625 \n",
       "L 9.28125 47.703125 \n",
       "L 2.6875 47.703125 \n",
       "L 2.6875 54.6875 \n",
       "L 9.28125 54.6875 \n",
       "L 9.28125 70.21875 \n",
       "z\n",
       "\" id=\"DejaVuSans-116\"/>\n",
       "      <path d=\"M 9.421875 54.6875 \n",
       "L 18.40625 54.6875 \n",
       "L 18.40625 0 \n",
       "L 9.421875 0 \n",
       "z\n",
       "M 9.421875 75.984375 \n",
       "L 18.40625 75.984375 \n",
       "L 18.40625 64.59375 \n",
       "L 9.421875 64.59375 \n",
       "z\n",
       "\" id=\"DejaVuSans-105\"/>\n",
       "      <path d=\"M 54.890625 33.015625 \n",
       "L 54.890625 0 \n",
       "L 45.90625 0 \n",
       "L 45.90625 32.71875 \n",
       "Q 45.90625 40.484375 42.875 44.328125 \n",
       "Q 39.84375 48.1875 33.796875 48.1875 \n",
       "Q 26.515625 48.1875 22.3125 43.546875 \n",
       "Q 18.109375 38.921875 18.109375 30.90625 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.1875 \n",
       "Q 21.34375 51.125 25.703125 53.5625 \n",
       "Q 30.078125 56 35.796875 56 \n",
       "Q 45.21875 56 50.046875 50.171875 \n",
       "Q 54.890625 44.34375 54.890625 33.015625 \n",
       "z\n",
       "\" id=\"DejaVuSans-110\"/>\n",
       "      <path d=\"M 50.984375 -16.609375 \n",
       "L 50.984375 -23.578125 \n",
       "L -0.984375 -23.578125 \n",
       "L -0.984375 -16.609375 \n",
       "z\n",
       "\" id=\"DejaVuSans-95\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(166.985938 53.154687)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-116\"/>\n",
       "      <use x=\"39.208984\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "      <use x=\"80.322266\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"141.601562\" xlink:href=\"#DejaVuSans-105\"/>\n",
       "      <use x=\"169.384766\" xlink:href=\"#DejaVuSans-110\"/>\n",
       "      <use x=\"232.763672\" xlink:href=\"#DejaVuSans-95\"/>\n",
       "      <use x=\"282.763672\" xlink:href=\"#DejaVuSans-101\"/>\n",
       "      <use x=\"344.287109\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "      <use x=\"385.384766\" xlink:href=\"#DejaVuSans-114\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"pd726c0490a\">\n",
       "   <rect height=\"135.9\" width=\"195.3\" x=\"33.2875\" y=\"7.2\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<Figure size 252x180 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "num_epochs = 25\n",
    "init_epoch = 0\n",
    "\n",
    "# Uncomment the following 2 lines if you wish to load a pre-trained/saved model\n",
    "# checkpoint_path = './ssd_outputs/model-29_0.1411931432526687.pth'   # Mention the model name to load \n",
    "# init_epoch = d2l.load(net, checkpoint_path, optimizer)\n",
    "\n",
    "animator = d2l.Animator(xlabel='epoch', xlim=[init_epoch+1, num_epochs],\n",
    "                        legend=['class error', 'bbox mae', 'train_err'])\n",
    "                        \n",
    "for epoch in range(init_epoch, num_epochs):\n",
    "        \n",
    "    net.train()\n",
    "    \n",
    "    train_loss = 0.0\n",
    "    loc_loss = 0.0\n",
    "    class_loss = 0.0\n",
    "\n",
    "    for i, (x, bb_true, class_true) in (enumerate(train_loader)):\n",
    "        \n",
    "        x = x.to(device)\n",
    "        bb_true = bb_true.to(device)\n",
    "        class_true = class_true.to(device)\n",
    "        \n",
    "        timer_start = time.time()\n",
    "        \n",
    "        anchors, cls_preds, bbox_preds = net(x)\n",
    "        \n",
    "        class_true = [*class_true.reshape((class_true.size(0), 1, 1))]\n",
    "        bb_true = [*bb_true.reshape((bb_true.size(0), 1, 1, 4))]\n",
    "        \n",
    "        bbox_preds = bbox_preds.reshape((-1, 5444, 4))\n",
    "        \n",
    "        # Label the category and offset of each anchor box\n",
    "        \n",
    "        anchors = anchors.to(device)\n",
    "                \n",
    "        batch_loss, batch_loc_loss, batch_class_loss = loss(cls_preds, bbox_preds, class_true, bb_true, anchors)\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        batch_loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        class_loss += batch_class_loss\n",
    "        loc_loss += batch_loc_loss\n",
    "        train_loss += batch_loss\n",
    "        \n",
    "#     Uncomment the following 2 line for saving the model every epoch\n",
    "#     path_to_checkpoints_dir = './ssd_outputs/'\n",
    "#     d2l.save(net, path_to_checkpoints_dir, epoch, optimizer, train_loss/len(train_loader))\n",
    "    \n",
    "    train_loss = (train_loss/len(train_loader)).detach().cpu().numpy()\n",
    "    loc_loss = (loc_loss/len(train_loader)).detach().cpu().numpy()\n",
    "    class_loss = (class_loss/len(train_loader)).detach().cpu().numpy()\n",
    "    \n",
    "#     print(class_loss, loc_loss, train_loss, epoch+1)\n",
    "    \n",
    "    # Uncomment the following if you wish to see the results after every epoch to see the learning effect\n",
    "    # Images will be saved to the 'results_every_epoch' directory\n",
    "    \n",
    "#     try:\n",
    "#         d2l.infer(net, epoch, 0.9, device)\n",
    "#     except Exception as e:\n",
    "#         print(e, 'error' + str(epoch+1))\n",
    "        \n",
    "    \n",
    "    animator.add(epoch, (class_loss, loc_loss, train_loss))  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Saving the final Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./ssd_outputs/model-24_0.19301878172775794.pth'"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path_to_checkpoints_dir = './ssd_outputs'\n",
    "d2l.save(net, path_to_checkpoints_dir, epoch, optimizer, train_loss/len(train_loader))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prediction\n",
    "\n",
    "In the prediction stage, we want to detect all objects of interest in the image. Below, we read the test image and transform its size. Then, we convert it to the four-dimensional format required by the convolutional layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "img = np.array(Image.open('../img/pikachu.jpg').convert('RGB').resize((256, 256), Image.BILINEAR))\n",
    "X = transforms.Compose([transforms.ToTensor()])(img).to(device)\n",
    "X = X.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We predict the bounding boxes based on the anchor boxes and their predicted offsets. Then, we use non-maximum suppression to remove similar bounding boxes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(X, nms_threshold):\n",
    "    background_threshold = 0.8\n",
    "    net.eval()\n",
    "    anchors, class_hat, bb_hat = net(X.unsqueeze(0))\n",
    "    anchors = anchors.to(device)\n",
    "    bb_hat = bb_hat.reshape((1, -1, 4))\n",
    "    bb_hat = d2l.invert_transformation(bb_hat.squeeze(0), anchors)\n",
    "    bb_hat = bb_hat * 256.0\n",
    "\n",
    "    class_hat = class_hat.sigmoid().squeeze(0)\n",
    "\n",
    "    bb_hat = bb_hat[class_hat[:,0] < background_threshold, :]\n",
    "\n",
    "\n",
    "    bb_hat = bb_hat.detach().cpu().numpy()\n",
    "    class_hat = class_hat[class_hat[:,0] < background_threshold, :]\n",
    "\n",
    "    class_preds = class_hat[:, 1:]\n",
    "\n",
    "    prob, class_id = torch.max(class_preds,1)\n",
    "\n",
    "    prob = prob.detach().cpu().numpy()\n",
    "    class_id = class_id.detach().cpu().numpy()\n",
    "\n",
    "\n",
    "\n",
    "    output_bb = [d2l.PredBoundingBox(probability=1 - prob[i],\n",
    "                                 class_id=class_id[i],\n",
    "                                 classname=id_cat[class_id[i]],\n",
    "                                 bounding_box=[bb_hat[i, 0], \n",
    "                                               bb_hat[i, 1], \n",
    "                                               bb_hat[i, 2], \n",
    "                                               bb_hat[i, 3]])\n",
    "                                 for i in range(0, len(prob))]\n",
    "\n",
    "    output_bb = sorted(output_bb, key = lambda x: x.probability, reverse=False)\n",
    "    \n",
    "    filtered_bb = d2l.non_max_suppression(output_bb, nms_threshold)\n",
    "    \n",
    "    return filtered_bb\n",
    "\n",
    "filtered_bb = predict(X, 0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we take all the bounding boxes and display them as the final output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (https://matplotlib.org/) -->\n",
       "<svg height=\"251.892656pt\" version=\"1.1\" viewBox=\"0 0 262.799687 251.892656\" width=\"262.799687pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 251.892656 \n",
       "L 262.799687 251.892656 \n",
       "L 262.799687 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 33.2875 228.014531 \n",
       "L 250.7275 228.014531 \n",
       "L 250.7275 10.574531 \n",
       "L 33.2875 10.574531 \n",
       "z\n",
       "\" style=\"fill:#ffffff;\"/>\n",
       "   </g>\n",
       "   <g clip-path=\"url(#p5ecd038ba4)\">\n",
       "    <image height=\"218\" id=\"image3acbf8c894\" transform=\"scale(1 -1)translate(0 -218)\" width=\"218\" x=\"33.2875\" xlink:href=\"data:image/png;base64,\n",
       "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\" y=\"-10.014531\"/>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_1\">\n",
       "    <g id=\"xtick_1\">\n",
       "     <g id=\"line2d_1\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L 0 3.5 \n",
       "\" id=\"m1a153c9435\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.712188\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_1\">\n",
       "      <!-- 0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 31.78125 66.40625 \n",
       "Q 24.171875 66.40625 20.328125 58.90625 \n",
       "Q 16.5 51.421875 16.5 36.375 \n",
       "Q 16.5 21.390625 20.328125 13.890625 \n",
       "Q 24.171875 6.390625 31.78125 6.390625 \n",
       "Q 39.453125 6.390625 43.28125 13.890625 \n",
       "Q 47.125 21.390625 47.125 36.375 \n",
       "Q 47.125 51.421875 43.28125 58.90625 \n",
       "Q 39.453125 66.40625 31.78125 66.40625 \n",
       "z\n",
       "M 31.78125 74.21875 \n",
       "Q 44.046875 74.21875 50.515625 64.515625 \n",
       "Q 56.984375 54.828125 56.984375 36.375 \n",
       "Q 56.984375 17.96875 50.515625 8.265625 \n",
       "Q 44.046875 -1.421875 31.78125 -1.421875 \n",
       "Q 19.53125 -1.421875 13.0625 8.265625 \n",
       "Q 6.59375 17.96875 6.59375 36.375 \n",
       "Q 6.59375 54.828125 13.0625 64.515625 \n",
       "Q 19.53125 74.21875 31.78125 74.21875 \n",
       "z\n",
       "\" id=\"DejaVuSans-48\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(30.530938 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_2\">\n",
       "     <g id=\"line2d_2\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"76.180937\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_2\">\n",
       "      <!-- 50 -->\n",
       "      <defs>\n",
       "       <path d=\"M 10.796875 72.90625 \n",
       "L 49.515625 72.90625 \n",
       "L 49.515625 64.59375 \n",
       "L 19.828125 64.59375 \n",
       "L 19.828125 46.734375 \n",
       "Q 21.96875 47.46875 24.109375 47.828125 \n",
       "Q 26.265625 48.1875 28.421875 48.1875 \n",
       "Q 40.625 48.1875 47.75 41.5 \n",
       "Q 54.890625 34.8125 54.890625 23.390625 \n",
       "Q 54.890625 11.625 47.5625 5.09375 \n",
       "Q 40.234375 -1.421875 26.90625 -1.421875 \n",
       "Q 22.3125 -1.421875 17.546875 -0.640625 \n",
       "Q 12.796875 0.140625 7.71875 1.703125 \n",
       "L 7.71875 11.625 \n",
       "Q 12.109375 9.234375 16.796875 8.0625 \n",
       "Q 21.484375 6.890625 26.703125 6.890625 \n",
       "Q 35.15625 6.890625 40.078125 11.328125 \n",
       "Q 45.015625 15.765625 45.015625 23.390625 \n",
       "Q 45.015625 31 40.078125 35.4375 \n",
       "Q 35.15625 39.890625 26.703125 39.890625 \n",
       "Q 22.75 39.890625 18.8125 39.015625 \n",
       "Q 14.890625 38.140625 10.796875 36.28125 \n",
       "z\n",
       "\" id=\"DejaVuSans-53\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(69.818438 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_3\">\n",
       "     <g id=\"line2d_3\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"118.649687\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_3\">\n",
       "      <!-- 100 -->\n",
       "      <defs>\n",
       "       <path d=\"M 12.40625 8.296875 \n",
       "L 28.515625 8.296875 \n",
       "L 28.515625 63.921875 \n",
       "L 10.984375 60.40625 \n",
       "L 10.984375 69.390625 \n",
       "L 28.421875 72.90625 \n",
       "L 38.28125 72.90625 \n",
       "L 38.28125 8.296875 \n",
       "L 54.390625 8.296875 \n",
       "L 54.390625 0 \n",
       "L 12.40625 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-49\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(109.105937 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_4\">\n",
       "     <g id=\"line2d_4\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"161.118437\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_4\">\n",
       "      <!-- 150 -->\n",
       "      <g transform=\"translate(151.574688 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_5\">\n",
       "     <g id=\"line2d_5\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"203.587187\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_5\">\n",
       "      <!-- 200 -->\n",
       "      <defs>\n",
       "       <path d=\"M 19.1875 8.296875 \n",
       "L 53.609375 8.296875 \n",
       "L 53.609375 0 \n",
       "L 7.328125 0 \n",
       "L 7.328125 8.296875 \n",
       "Q 12.9375 14.109375 22.625 23.890625 \n",
       "Q 32.328125 33.6875 34.8125 36.53125 \n",
       "Q 39.546875 41.84375 41.421875 45.53125 \n",
       "Q 43.3125 49.21875 43.3125 52.78125 \n",
       "Q 43.3125 58.59375 39.234375 62.25 \n",
       "Q 35.15625 65.921875 28.609375 65.921875 \n",
       "Q 23.96875 65.921875 18.8125 64.3125 \n",
       "Q 13.671875 62.703125 7.8125 59.421875 \n",
       "L 7.8125 69.390625 \n",
       "Q 13.765625 71.78125 18.9375 73 \n",
       "Q 24.125 74.21875 28.421875 74.21875 \n",
       "Q 39.75 74.21875 46.484375 68.546875 \n",
       "Q 53.21875 62.890625 53.21875 53.421875 \n",
       "Q 53.21875 48.921875 51.53125 44.890625 \n",
       "Q 49.859375 40.875 45.40625 35.40625 \n",
       "Q 44.1875 33.984375 37.640625 27.21875 \n",
       "Q 31.109375 20.453125 19.1875 8.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-50\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(194.043438 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_6\">\n",
       "     <g id=\"line2d_6\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"246.055937\" xlink:href=\"#m1a153c9435\" y=\"228.014531\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_6\">\n",
       "      <!-- 250 -->\n",
       "      <g transform=\"translate(236.512188 242.612969)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_2\">\n",
       "    <g id=\"ytick_1\">\n",
       "     <g id=\"line2d_7\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L -3.5 0 \n",
       "\" id=\"madaa8244af\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"10.999219\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_7\">\n",
       "      <!-- 0 -->\n",
       "      <g transform=\"translate(19.925 14.798438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_2\">\n",
       "     <g id=\"line2d_8\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"53.467969\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_8\">\n",
       "      <!-- 50 -->\n",
       "      <g transform=\"translate(13.5625 57.267188)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_3\">\n",
       "     <g id=\"line2d_9\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"95.936719\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_9\">\n",
       "      <!-- 100 -->\n",
       "      <g transform=\"translate(7.2 99.735938)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_4\">\n",
       "     <g id=\"line2d_10\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"138.405469\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_10\">\n",
       "      <!-- 150 -->\n",
       "      <g transform=\"translate(7.2 142.204688)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_5\">\n",
       "     <g id=\"line2d_11\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"180.874219\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_11\">\n",
       "      <!-- 200 -->\n",
       "      <g transform=\"translate(7.2 184.673438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_6\">\n",
       "     <g id=\"line2d_12\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"33.2875\" xlink:href=\"#madaa8244af\" y=\"223.342969\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_12\">\n",
       "      <!-- 250 -->\n",
       "      <g transform=\"translate(7.2 227.142188)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "       <use x=\"127.246094\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 33.2875 228.014531 \n",
       "L 33.2875 10.574531 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_4\">\n",
       "    <path d=\"M 250.7275 228.014531 \n",
       "L 250.7275 10.574531 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_5\">\n",
       "    <path d=\"M 33.2875 228.014531 \n",
       "L 250.7275 228.014531 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_6\">\n",
       "    <path d=\"M 33.2875 10.574531 \n",
       "L 250.7275 10.574531 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"p5ecd038ba4\">\n",
       "   <rect height=\"217.44\" width=\"217.44\" x=\"33.2875\" y=\"10.574531\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def display(img, output):\n",
    "\n",
    "    img = d2l.draw_boxes(img, [bb.bounding_box for bb in filtered_bb])\n",
    "    img = d2l.draw_text(img, [str(bb.probability)[:5] for bb in filtered_bb], [bb.bounding_box for bb in filtered_bb])\n",
    "    d2l.plt.imshow(img)\n",
    "    d2l.plt.show()\n",
    "\n",
    "display(img, filtered_bb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Summary\n",
    "\n",
    "* SSD is a multiscale object detection model. This model generates different numbers of anchor boxes of different sizes based on the base network block and each multiscale feature block and predicts the categories and offsets of the anchor boxes to detect objects of different sizes.\n",
    "* During SSD model training, the loss function is calculated using the predicted and labeled category and offset values.\n",
    "\n",
    "\n",
    "\n",
    "## Exercises\n",
    "\n",
    "* Due to space limitations, we have ignored some of the implementation details of SSD models in this experiment. Can you further improve the model in the following areas?\n",
    "\n",
    "\n",
    "### Loss Function\n",
    "\n",
    "For the predicted offsets, replace $L_1$ norm loss with $L_1$ regularization loss. This loss function uses a square function around zero for greater smoothness. This is the regularized area controlled by the hyper-parameter $\\sigma$:\n",
    "\n",
    "$$\n",
    "f(x) =\n",
    "    \\begin{cases}\n",
    "    (\\sigma x)^2/2,& \\text{if }|x| < 1/\\sigma^2\\\\\n",
    "    |x|-0.5/\\sigma^2,& \\text{otherwise}\n",
    "    \\end{cases}\n",
    "$$\n",
    "\n",
    "When $\\sigma$ is large, this loss is similar to the $L_1$ norm loss. When the value is small, the loss function is smoother."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "sigmas = [10, 1, 0.5]\n",
    "lines = ['-', '--', '-.']\n",
    "x = np.arange(-2, 2, 0.1)\n",
    "d2l.set_figsize()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "def smooth_l1(x, scalar):\n",
    "    a = []\n",
    "    for i in x:\n",
    "                \n",
    "        if abs(i) < 1/((scalar)**2):\n",
    "            a.append(((scalar*i)**2)/2)\n",
    "        \n",
    "        else:\n",
    "            a.append(abs(i) - 0.5/((scalar)**2))\n",
    "    \n",
    "    return np.array(a)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (https://matplotlib.org/) -->\n",
       "<svg height=\"166.978125pt\" version=\"1.1\" viewBox=\"0 0 232.603125 166.978125\" width=\"232.603125pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 166.978125 \n",
       "L 232.603125 166.978125 \n",
       "L 232.603125 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 30.103125 143.1 \n",
       "L 225.403125 143.1 \n",
       "L 225.403125 7.2 \n",
       "L 30.103125 7.2 \n",
       "z\n",
       "\" style=\"fill:#ffffff;\"/>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_1\">\n",
       "    <g id=\"xtick_1\">\n",
       "     <g id=\"line2d_1\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L 0 3.5 \n",
       "\" id=\"m3246c99a72\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"38.980398\" xlink:href=\"#m3246c99a72\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_1\">\n",
       "      <!-- −2 -->\n",
       "      <defs>\n",
       "       <path d=\"M 10.59375 35.5 \n",
       "L 73.1875 35.5 \n",
       "L 73.1875 27.203125 \n",
       "L 10.59375 27.203125 \n",
       "z\n",
       "\" id=\"DejaVuSans-8722\"/>\n",
       "       <path d=\"M 19.1875 8.296875 \n",
       "L 53.609375 8.296875 \n",
       "L 53.609375 0 \n",
       "L 7.328125 0 \n",
       "L 7.328125 8.296875 \n",
       "Q 12.9375 14.109375 22.625 23.890625 \n",
       "Q 32.328125 33.6875 34.8125 36.53125 \n",
       "Q 39.546875 41.84375 41.421875 45.53125 \n",
       "Q 43.3125 49.21875 43.3125 52.78125 \n",
       "Q 43.3125 58.59375 39.234375 62.25 \n",
       "Q 35.15625 65.921875 28.609375 65.921875 \n",
       "Q 23.96875 65.921875 18.8125 64.3125 \n",
       "Q 13.671875 62.703125 7.8125 59.421875 \n",
       "L 7.8125 69.390625 \n",
       "Q 13.765625 71.78125 18.9375 73 \n",
       "Q 24.125 74.21875 28.421875 74.21875 \n",
       "Q 39.75 74.21875 46.484375 68.546875 \n",
       "Q 53.21875 62.890625 53.21875 53.421875 \n",
       "Q 53.21875 48.921875 51.53125 44.890625 \n",
       "Q 49.859375 40.875 45.40625 35.40625 \n",
       "Q 44.1875 33.984375 37.640625 27.21875 \n",
       "Q 31.109375 20.453125 19.1875 8.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-50\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(31.609304 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-8722\"/>\n",
       "       <use x=\"83.789062\" xlink:href=\"#DejaVuSans-50\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_2\">\n",
       "     <g id=\"line2d_2\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"84.504873\" xlink:href=\"#m3246c99a72\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_2\">\n",
       "      <!-- −1 -->\n",
       "      <defs>\n",
       "       <path d=\"M 12.40625 8.296875 \n",
       "L 28.515625 8.296875 \n",
       "L 28.515625 63.921875 \n",
       "L 10.984375 60.40625 \n",
       "L 10.984375 69.390625 \n",
       "L 28.421875 72.90625 \n",
       "L 38.28125 72.90625 \n",
       "L 38.28125 8.296875 \n",
       "L 54.390625 8.296875 \n",
       "L 54.390625 0 \n",
       "L 12.40625 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-49\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(77.13378 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-8722\"/>\n",
       "       <use x=\"83.789062\" xlink:href=\"#DejaVuSans-49\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_3\">\n",
       "     <g id=\"line2d_3\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"130.029349\" xlink:href=\"#m3246c99a72\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_3\">\n",
       "      <!-- 0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 31.78125 66.40625 \n",
       "Q 24.171875 66.40625 20.328125 58.90625 \n",
       "Q 16.5 51.421875 16.5 36.375 \n",
       "Q 16.5 21.390625 20.328125 13.890625 \n",
       "Q 24.171875 6.390625 31.78125 6.390625 \n",
       "Q 39.453125 6.390625 43.28125 13.890625 \n",
       "Q 47.125 21.390625 47.125 36.375 \n",
       "Q 47.125 51.421875 43.28125 58.90625 \n",
       "Q 39.453125 66.40625 31.78125 66.40625 \n",
       "z\n",
       "M 31.78125 74.21875 \n",
       "Q 44.046875 74.21875 50.515625 64.515625 \n",
       "Q 56.984375 54.828125 56.984375 36.375 \n",
       "Q 56.984375 17.96875 50.515625 8.265625 \n",
       "Q 44.046875 -1.421875 31.78125 -1.421875 \n",
       "Q 19.53125 -1.421875 13.0625 8.265625 \n",
       "Q 6.59375 17.96875 6.59375 36.375 \n",
       "Q 6.59375 54.828125 13.0625 64.515625 \n",
       "Q 19.53125 74.21875 31.78125 74.21875 \n",
       "z\n",
       "\" id=\"DejaVuSans-48\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(126.848099 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_4\">\n",
       "     <g id=\"line2d_4\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"175.553824\" xlink:href=\"#m3246c99a72\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_4\">\n",
       "      <!-- 1 -->\n",
       "      <g transform=\"translate(172.372574 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_5\">\n",
       "     <g id=\"line2d_5\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"221.0783\" xlink:href=\"#m3246c99a72\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_5\">\n",
       "      <!-- 2 -->\n",
       "      <g transform=\"translate(217.89705 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_2\">\n",
       "    <g id=\"ytick_1\">\n",
       "     <g id=\"line2d_6\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L -3.5 0 \n",
       "\" id=\"mc7c8ced16c\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"30.103125\" xlink:href=\"#mc7c8ced16c\" y=\"136.922727\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_6\">\n",
       "      <!-- 0.0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 10.6875 12.40625 \n",
       "L 21 12.40625 \n",
       "L 21 0 \n",
       "L 10.6875 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-46\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(7.2 140.721946)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_2\">\n",
       "     <g id=\"line2d_7\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"30.103125\" xlink:href=\"#mc7c8ced16c\" y=\"105.958954\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_7\">\n",
       "      <!-- 0.5 -->\n",
       "      <defs>\n",
       "       <path d=\"M 10.796875 72.90625 \n",
       "L 49.515625 72.90625 \n",
       "L 49.515625 64.59375 \n",
       "L 19.828125 64.59375 \n",
       "L 19.828125 46.734375 \n",
       "Q 21.96875 47.46875 24.109375 47.828125 \n",
       "Q 26.265625 48.1875 28.421875 48.1875 \n",
       "Q 40.625 48.1875 47.75 41.5 \n",
       "Q 54.890625 34.8125 54.890625 23.390625 \n",
       "Q 54.890625 11.625 47.5625 5.09375 \n",
       "Q 40.234375 -1.421875 26.90625 -1.421875 \n",
       "Q 22.3125 -1.421875 17.546875 -0.640625 \n",
       "Q 12.796875 0.140625 7.71875 1.703125 \n",
       "L 7.71875 11.625 \n",
       "Q 12.109375 9.234375 16.796875 8.0625 \n",
       "Q 21.484375 6.890625 26.703125 6.890625 \n",
       "Q 35.15625 6.890625 40.078125 11.328125 \n",
       "Q 45.015625 15.765625 45.015625 23.390625 \n",
       "Q 45.015625 31 40.078125 35.4375 \n",
       "Q 35.15625 39.890625 26.703125 39.890625 \n",
       "Q 22.75 39.890625 18.8125 39.015625 \n",
       "Q 14.890625 38.140625 10.796875 36.28125 \n",
       "z\n",
       "\" id=\"DejaVuSans-53\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(7.2 109.758173)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_3\">\n",
       "     <g id=\"line2d_8\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"30.103125\" xlink:href=\"#mc7c8ced16c\" y=\"74.995181\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_8\">\n",
       "      <!-- 1.0 -->\n",
       "      <g transform=\"translate(7.2 78.7944)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_4\">\n",
       "     <g id=\"line2d_9\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"30.103125\" xlink:href=\"#mc7c8ced16c\" y=\"44.031408\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_9\">\n",
       "      <!-- 1.5 -->\n",
       "      <g transform=\"translate(7.2 47.830627)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_5\">\n",
       "     <g id=\"line2d_10\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"30.103125\" xlink:href=\"#mc7c8ced16c\" y=\"13.067635\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_10\">\n",
       "      <!-- 2.0 -->\n",
       "      <g transform=\"translate(7.2 16.866854)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"line2d_11\">\n",
       "    <path clip-path=\"url(#pcfcd477b19)\" d=\"M 38.980398 13.377273 \n",
       "L 43.532845 19.570027 \n",
       "L 48.085293 25.762782 \n",
       "L 52.63774 31.955537 \n",
       "L 57.190188 38.148291 \n",
       "L 61.742635 44.341046 \n",
       "L 66.295083 50.5338 \n",
       "L 70.847531 56.726555 \n",
       "L 75.399978 62.91931 \n",
       "L 79.952426 69.112064 \n",
       "L 84.504873 75.304819 \n",
       "L 89.057321 81.497573 \n",
       "L 93.609768 87.690328 \n",
       "L 98.162216 93.883083 \n",
       "L 102.714663 100.075837 \n",
       "L 107.267111 106.268592 \n",
       "L 111.819559 112.461347 \n",
       "L 116.372006 118.654101 \n",
       "L 120.924454 124.846856 \n",
       "L 125.476901 131.03961 \n",
       "L 130.029349 136.922727 \n",
       "L 134.581796 131.03961 \n",
       "L 139.134244 124.846856 \n",
       "L 143.686691 118.654101 \n",
       "L 148.239139 112.461347 \n",
       "L 152.791587 106.268592 \n",
       "L 157.344034 100.075837 \n",
       "L 161.896482 93.883083 \n",
       "L 166.448929 87.690328 \n",
       "L 171.001377 81.497573 \n",
       "L 175.553824 75.304819 \n",
       "L 180.106272 69.112064 \n",
       "L 184.658719 62.91931 \n",
       "L 189.211167 56.726555 \n",
       "L 193.763615 50.5338 \n",
       "L 198.316062 44.341046 \n",
       "L 202.86851 38.148291 \n",
       "L 207.420957 31.955537 \n",
       "L 211.973405 25.762782 \n",
       "L 216.525852 19.570027 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_12\">\n",
       "    <path clip-path=\"url(#pcfcd477b19)\" d=\"M 38.980398 44.031408 \n",
       "L 43.532845 50.224163 \n",
       "L 48.085293 56.416917 \n",
       "L 52.63774 62.609672 \n",
       "L 57.190188 68.802427 \n",
       "L 61.742635 74.995181 \n",
       "L 66.295083 81.187936 \n",
       "L 70.847531 87.38069 \n",
       "L 75.399978 93.573445 \n",
       "L 79.952426 99.7662 \n",
       "L 84.504873 105.958954 \n",
       "L 89.057321 111.842071 \n",
       "L 93.609768 117.105913 \n",
       "L 98.162216 121.750478 \n",
       "L 102.714663 125.775769 \n",
       "L 107.267111 129.181784 \n",
       "L 111.819559 131.968524 \n",
       "L 116.372006 134.135988 \n",
       "L 120.924454 135.684176 \n",
       "L 125.476901 136.61309 \n",
       "L 130.029349 136.922727 \n",
       "L 134.581796 136.61309 \n",
       "L 139.134244 135.684176 \n",
       "L 143.686691 134.135988 \n",
       "L 148.239139 131.968524 \n",
       "L 152.791587 129.181784 \n",
       "L 157.344034 125.775769 \n",
       "L 161.896482 121.750478 \n",
       "L 166.448929 117.105913 \n",
       "L 171.001377 111.842071 \n",
       "L 175.553824 105.958954 \n",
       "L 180.106272 99.7662 \n",
       "L 184.658719 93.573445 \n",
       "L 189.211167 87.38069 \n",
       "L 193.763615 81.187936 \n",
       "L 198.316062 74.995181 \n",
       "L 202.86851 68.802427 \n",
       "L 207.420957 62.609672 \n",
       "L 211.973405 56.416917 \n",
       "L 216.525852 50.224163 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-dasharray:5.55,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_13\">\n",
       "    <path clip-path=\"url(#pcfcd477b19)\" d=\"M 38.980398 105.958954 \n",
       "L 43.532845 108.977922 \n",
       "L 48.085293 111.842071 \n",
       "L 52.63774 114.551401 \n",
       "L 57.190188 117.105913 \n",
       "L 61.742635 119.505605 \n",
       "L 66.295083 121.750478 \n",
       "L 70.847531 123.840533 \n",
       "L 75.399978 125.775769 \n",
       "L 79.952426 127.556186 \n",
       "L 84.504873 129.181784 \n",
       "L 89.057321 130.652563 \n",
       "L 93.609768 131.968524 \n",
       "L 98.162216 133.129665 \n",
       "L 102.714663 134.135988 \n",
       "L 107.267111 134.987491 \n",
       "L 111.819559 135.684176 \n",
       "L 116.372006 136.226042 \n",
       "L 120.924454 136.61309 \n",
       "L 125.476901 136.845318 \n",
       "L 130.029349 136.922727 \n",
       "L 134.581796 136.845318 \n",
       "L 139.134244 136.61309 \n",
       "L 143.686691 136.226042 \n",
       "L 148.239139 135.684176 \n",
       "L 152.791587 134.987491 \n",
       "L 157.344034 134.135988 \n",
       "L 161.896482 133.129665 \n",
       "L 166.448929 131.968524 \n",
       "L 171.001377 130.652563 \n",
       "L 175.553824 129.181784 \n",
       "L 180.106272 127.556186 \n",
       "L 184.658719 125.775769 \n",
       "L 189.211167 123.840533 \n",
       "L 193.763615 121.750478 \n",
       "L 198.316062 119.505605 \n",
       "L 202.86851 117.105913 \n",
       "L 207.420957 114.551401 \n",
       "L 211.973405 111.842071 \n",
       "L 216.525852 108.977922 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-dasharray:9.6,2.4,1.5,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 30.103125 143.1 \n",
       "L 30.103125 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_4\">\n",
       "    <path d=\"M 225.403125 143.1 \n",
       "L 225.403125 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_5\">\n",
       "    <path d=\"M 30.103125 143.1 \n",
       "L 225.403125 143.1 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_6\">\n",
       "    <path d=\"M 30.103125 7.2 \n",
       "L 225.403125 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"legend_1\">\n",
       "    <g id=\"patch_7\">\n",
       "     <path d=\"M 81.328125 59.234375 \n",
       "L 174.178125 59.234375 \n",
       "Q 176.178125 59.234375 176.178125 57.234375 \n",
       "L 176.178125 14.2 \n",
       "Q 176.178125 12.2 174.178125 12.2 \n",
       "L 81.328125 12.2 \n",
       "Q 79.328125 12.2 79.328125 14.2 \n",
       "L 79.328125 57.234375 \n",
       "Q 79.328125 59.234375 81.328125 59.234375 \n",
       "z\n",
       "\" style=\"fill:#ffffff;opacity:0.8;stroke:#cccccc;stroke-linejoin:miter;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_14\">\n",
       "     <path d=\"M 83.328125 20.298438 \n",
       "L 103.328125 20.298438 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_15\"/>\n",
       "    <g id=\"text_11\">\n",
       "     <!-- sigma=10.0 -->\n",
       "     <defs>\n",
       "      <path d=\"M 44.28125 53.078125 \n",
       "L 44.28125 44.578125 \n",
       "Q 40.484375 46.53125 36.375 47.5 \n",
       "Q 32.28125 48.484375 27.875 48.484375 \n",
       "Q 21.1875 48.484375 17.84375 46.4375 \n",
       "Q 14.5 44.390625 14.5 40.28125 \n",
       "Q 14.5 37.15625 16.890625 35.375 \n",
       "Q 19.28125 33.59375 26.515625 31.984375 \n",
       "L 29.59375 31.296875 \n",
       "Q 39.15625 29.25 43.1875 25.515625 \n",
       "Q 47.21875 21.78125 47.21875 15.09375 \n",
       "Q 47.21875 7.46875 41.1875 3.015625 \n",
       "Q 35.15625 -1.421875 24.609375 -1.421875 \n",
       "Q 20.21875 -1.421875 15.453125 -0.5625 \n",
       "Q 10.6875 0.296875 5.421875 2 \n",
       "L 5.421875 11.28125 \n",
       "Q 10.40625 8.6875 15.234375 7.390625 \n",
       "Q 20.0625 6.109375 24.8125 6.109375 \n",
       "Q 31.15625 6.109375 34.5625 8.28125 \n",
       "Q 37.984375 10.453125 37.984375 14.40625 \n",
       "Q 37.984375 18.0625 35.515625 20.015625 \n",
       "Q 33.0625 21.96875 24.703125 23.78125 \n",
       "L 21.578125 24.515625 \n",
       "Q 13.234375 26.265625 9.515625 29.90625 \n",
       "Q 5.8125 33.546875 5.8125 39.890625 \n",
       "Q 5.8125 47.609375 11.28125 51.796875 \n",
       "Q 16.75 56 26.8125 56 \n",
       "Q 31.78125 56 36.171875 55.265625 \n",
       "Q 40.578125 54.546875 44.28125 53.078125 \n",
       "z\n",
       "\" id=\"DejaVuSans-115\"/>\n",
       "      <path d=\"M 9.421875 54.6875 \n",
       "L 18.40625 54.6875 \n",
       "L 18.40625 0 \n",
       "L 9.421875 0 \n",
       "z\n",
       "M 9.421875 75.984375 \n",
       "L 18.40625 75.984375 \n",
       "L 18.40625 64.59375 \n",
       "L 9.421875 64.59375 \n",
       "z\n",
       "\" id=\"DejaVuSans-105\"/>\n",
       "      <path d=\"M 45.40625 27.984375 \n",
       "Q 45.40625 37.75 41.375 43.109375 \n",
       "Q 37.359375 48.484375 30.078125 48.484375 \n",
       "Q 22.859375 48.484375 18.828125 43.109375 \n",
       "Q 14.796875 37.75 14.796875 27.984375 \n",
       "Q 14.796875 18.265625 18.828125 12.890625 \n",
       "Q 22.859375 7.515625 30.078125 7.515625 \n",
       "Q 37.359375 7.515625 41.375 12.890625 \n",
       "Q 45.40625 18.265625 45.40625 27.984375 \n",
       "z\n",
       "M 54.390625 6.78125 \n",
       "Q 54.390625 -7.171875 48.1875 -13.984375 \n",
       "Q 42 -20.796875 29.203125 -20.796875 \n",
       "Q 24.46875 -20.796875 20.265625 -20.09375 \n",
       "Q 16.0625 -19.390625 12.109375 -17.921875 \n",
       "L 12.109375 -9.1875 \n",
       "Q 16.0625 -11.328125 19.921875 -12.34375 \n",
       "Q 23.78125 -13.375 27.78125 -13.375 \n",
       "Q 36.625 -13.375 41.015625 -8.765625 \n",
       "Q 45.40625 -4.15625 45.40625 5.171875 \n",
       "L 45.40625 9.625 \n",
       "Q 42.625 4.78125 38.28125 2.390625 \n",
       "Q 33.9375 0 27.875 0 \n",
       "Q 17.828125 0 11.671875 7.65625 \n",
       "Q 5.515625 15.328125 5.515625 27.984375 \n",
       "Q 5.515625 40.671875 11.671875 48.328125 \n",
       "Q 17.828125 56 27.875 56 \n",
       "Q 33.9375 56 38.28125 53.609375 \n",
       "Q 42.625 51.21875 45.40625 46.390625 \n",
       "L 45.40625 54.6875 \n",
       "L 54.390625 54.6875 \n",
       "z\n",
       "\" id=\"DejaVuSans-103\"/>\n",
       "      <path d=\"M 52 44.1875 \n",
       "Q 55.375 50.25 60.0625 53.125 \n",
       "Q 64.75 56 71.09375 56 \n",
       "Q 79.640625 56 84.28125 50.015625 \n",
       "Q 88.921875 44.046875 88.921875 33.015625 \n",
       "L 88.921875 0 \n",
       "L 79.890625 0 \n",
       "L 79.890625 32.71875 \n",
       "Q 79.890625 40.578125 77.09375 44.375 \n",
       "Q 74.3125 48.1875 68.609375 48.1875 \n",
       "Q 61.625 48.1875 57.5625 43.546875 \n",
       "Q 53.515625 38.921875 53.515625 30.90625 \n",
       "L 53.515625 0 \n",
       "L 44.484375 0 \n",
       "L 44.484375 32.71875 \n",
       "Q 44.484375 40.625 41.703125 44.40625 \n",
       "Q 38.921875 48.1875 33.109375 48.1875 \n",
       "Q 26.21875 48.1875 22.15625 43.53125 \n",
       "Q 18.109375 38.875 18.109375 30.90625 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.1875 \n",
       "Q 21.1875 51.21875 25.484375 53.609375 \n",
       "Q 29.78125 56 35.6875 56 \n",
       "Q 41.65625 56 45.828125 52.96875 \n",
       "Q 50 49.953125 52 44.1875 \n",
       "z\n",
       "\" id=\"DejaVuSans-109\"/>\n",
       "      <path d=\"M 34.28125 27.484375 \n",
       "Q 23.390625 27.484375 19.1875 25 \n",
       "Q 14.984375 22.515625 14.984375 16.5 \n",
       "Q 14.984375 11.71875 18.140625 8.90625 \n",
       "Q 21.296875 6.109375 26.703125 6.109375 \n",
       "Q 34.1875 6.109375 38.703125 11.40625 \n",
       "Q 43.21875 16.703125 43.21875 25.484375 \n",
       "L 43.21875 27.484375 \n",
       "z\n",
       "M 52.203125 31.203125 \n",
       "L 52.203125 0 \n",
       "L 43.21875 0 \n",
       "L 43.21875 8.296875 \n",
       "Q 40.140625 3.328125 35.546875 0.953125 \n",
       "Q 30.953125 -1.421875 24.3125 -1.421875 \n",
       "Q 15.921875 -1.421875 10.953125 3.296875 \n",
       "Q 6 8.015625 6 15.921875 \n",
       "Q 6 25.140625 12.171875 29.828125 \n",
       "Q 18.359375 34.515625 30.609375 34.515625 \n",
       "L 43.21875 34.515625 \n",
       "L 43.21875 35.40625 \n",
       "Q 43.21875 41.609375 39.140625 45 \n",
       "Q 35.0625 48.390625 27.6875 48.390625 \n",
       "Q 23 48.390625 18.546875 47.265625 \n",
       "Q 14.109375 46.140625 10.015625 43.890625 \n",
       "L 10.015625 52.203125 \n",
       "Q 14.9375 54.109375 19.578125 55.046875 \n",
       "Q 24.21875 56 28.609375 56 \n",
       "Q 40.484375 56 46.34375 49.84375 \n",
       "Q 52.203125 43.703125 52.203125 31.203125 \n",
       "z\n",
       "\" id=\"DejaVuSans-97\"/>\n",
       "      <path d=\"M 10.59375 45.40625 \n",
       "L 73.1875 45.40625 \n",
       "L 73.1875 37.203125 \n",
       "L 10.59375 37.203125 \n",
       "z\n",
       "M 10.59375 25.484375 \n",
       "L 73.1875 25.484375 \n",
       "L 73.1875 17.1875 \n",
       "L 10.59375 17.1875 \n",
       "z\n",
       "\" id=\"DejaVuSans-61\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(111.328125 23.798438)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-115\"/>\n",
       "      <use x=\"52.099609\" xlink:href=\"#DejaVuSans-105\"/>\n",
       "      <use x=\"79.882812\" xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"143.359375\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"240.771484\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"302.050781\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"385.839844\" xlink:href=\"#DejaVuSans-49\"/>\n",
       "      <use x=\"449.462891\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      <use x=\"513.085938\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"544.873047\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_16\">\n",
       "     <path d=\"M 83.328125 34.976562 \n",
       "L 103.328125 34.976562 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-dasharray:5.55,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_17\"/>\n",
       "    <g id=\"text_12\">\n",
       "     <!-- sigma=1.0 -->\n",
       "     <g transform=\"translate(111.328125 38.476562)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-115\"/>\n",
       "      <use x=\"52.099609\" xlink:href=\"#DejaVuSans-105\"/>\n",
       "      <use x=\"79.882812\" xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"143.359375\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"240.771484\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"302.050781\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"385.839844\" xlink:href=\"#DejaVuSans-49\"/>\n",
       "      <use x=\"449.462891\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"481.25\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_18\">\n",
       "     <path d=\"M 83.328125 49.654688 \n",
       "L 103.328125 49.654688 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-dasharray:9.6,2.4,1.5,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_19\"/>\n",
       "    <g id=\"text_13\">\n",
       "     <!-- sigma=0.5 -->\n",
       "     <g transform=\"translate(111.328125 53.154688)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-115\"/>\n",
       "      <use x=\"52.099609\" xlink:href=\"#DejaVuSans-105\"/>\n",
       "      <use x=\"79.882812\" xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"143.359375\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"240.771484\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"302.050781\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"385.839844\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      <use x=\"449.462891\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"481.25\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"pcfcd477b19\">\n",
       "   <rect height=\"135.9\" width=\"195.3\" x=\"30.103125\" y=\"7.2\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<Figure size 252x180 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a1 = []\n",
    "for l, s in zip(lines, sigmas):\n",
    "    y = smooth_l1(x, scalar=s)\n",
    "    d2l.plt.plot(x, y, l, label='sigma=%.1f' % s)\n",
    "    \n",
    "d2l.plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the experiment, we used cross-entropy loss for category prediction. Now,\n",
    "assume that the prediction probability of the actual category $j$ is $p_j$ and\n",
    "the cross-entropy loss is $-\\log p_j$. We can also use the focal loss\n",
    ":cite:`Lin.Goyal.Girshick.ea.2017`. Given the positive hyper-parameters $\\gamma$\n",
    "and $\\alpha$, this loss is defined as:\n",
    "\n",
    "$$ - \\alpha (1-p_j)^{\\gamma} \\log p_j.$$\n",
    "\n",
    "As you can see, by increasing $\\gamma$, we can effectively reduce the loss when the probability of predicting the correct category is high."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (https://matplotlib.org/) -->\n",
       "<svg height=\"166.978125pt\" version=\"1.1\" viewBox=\"0 0 223.948478 166.978125\" width=\"223.948478pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 166.978125 \n",
       "L 223.948478 166.978125 \n",
       "L 223.948478 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 20.5625 143.1 \n",
       "L 215.8625 143.1 \n",
       "L 215.8625 7.2 \n",
       "L 20.5625 7.2 \n",
       "z\n",
       "\" style=\"fill:#ffffff;\"/>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_1\">\n",
       "    <g id=\"xtick_1\">\n",
       "     <g id=\"line2d_1\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L 0 3.5 \n",
       "\" id=\"m805066baff\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"27.628084\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_1\">\n",
       "      <!-- 0.0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 31.78125 66.40625 \n",
       "Q 24.171875 66.40625 20.328125 58.90625 \n",
       "Q 16.5 51.421875 16.5 36.375 \n",
       "Q 16.5 21.390625 20.328125 13.890625 \n",
       "Q 24.171875 6.390625 31.78125 6.390625 \n",
       "Q 39.453125 6.390625 43.28125 13.890625 \n",
       "Q 47.125 21.390625 47.125 36.375 \n",
       "Q 47.125 51.421875 43.28125 58.90625 \n",
       "Q 39.453125 66.40625 31.78125 66.40625 \n",
       "z\n",
       "M 31.78125 74.21875 \n",
       "Q 44.046875 74.21875 50.515625 64.515625 \n",
       "Q 56.984375 54.828125 56.984375 36.375 \n",
       "Q 56.984375 17.96875 50.515625 8.265625 \n",
       "Q 44.046875 -1.421875 31.78125 -1.421875 \n",
       "Q 19.53125 -1.421875 13.0625 8.265625 \n",
       "Q 6.59375 17.96875 6.59375 36.375 \n",
       "Q 6.59375 54.828125 13.0625 64.515625 \n",
       "Q 19.53125 74.21875 31.78125 74.21875 \n",
       "z\n",
       "\" id=\"DejaVuSans-48\"/>\n",
       "       <path d=\"M 10.6875 12.40625 \n",
       "L 21 12.40625 \n",
       "L 21 0 \n",
       "L 10.6875 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-46\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(19.676522 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_2\">\n",
       "     <g id=\"line2d_2\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"63.861851\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_2\">\n",
       "      <!-- 0.2 -->\n",
       "      <defs>\n",
       "       <path d=\"M 19.1875 8.296875 \n",
       "L 53.609375 8.296875 \n",
       "L 53.609375 0 \n",
       "L 7.328125 0 \n",
       "L 7.328125 8.296875 \n",
       "Q 12.9375 14.109375 22.625 23.890625 \n",
       "Q 32.328125 33.6875 34.8125 36.53125 \n",
       "Q 39.546875 41.84375 41.421875 45.53125 \n",
       "Q 43.3125 49.21875 43.3125 52.78125 \n",
       "Q 43.3125 58.59375 39.234375 62.25 \n",
       "Q 35.15625 65.921875 28.609375 65.921875 \n",
       "Q 23.96875 65.921875 18.8125 64.3125 \n",
       "Q 13.671875 62.703125 7.8125 59.421875 \n",
       "L 7.8125 69.390625 \n",
       "Q 13.765625 71.78125 18.9375 73 \n",
       "Q 24.125 74.21875 28.421875 74.21875 \n",
       "Q 39.75 74.21875 46.484375 68.546875 \n",
       "Q 53.21875 62.890625 53.21875 53.421875 \n",
       "Q 53.21875 48.921875 51.53125 44.890625 \n",
       "Q 49.859375 40.875 45.40625 35.40625 \n",
       "Q 44.1875 33.984375 37.640625 27.21875 \n",
       "Q 31.109375 20.453125 19.1875 8.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-50\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(55.910288 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-50\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_3\">\n",
       "     <g id=\"line2d_3\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"100.095617\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_3\">\n",
       "      <!-- 0.4 -->\n",
       "      <defs>\n",
       "       <path d=\"M 37.796875 64.3125 \n",
       "L 12.890625 25.390625 \n",
       "L 37.796875 25.390625 \n",
       "z\n",
       "M 35.203125 72.90625 \n",
       "L 47.609375 72.90625 \n",
       "L 47.609375 25.390625 \n",
       "L 58.015625 25.390625 \n",
       "L 58.015625 17.1875 \n",
       "L 47.609375 17.1875 \n",
       "L 47.609375 0 \n",
       "L 37.796875 0 \n",
       "L 37.796875 17.1875 \n",
       "L 4.890625 17.1875 \n",
       "L 4.890625 26.703125 \n",
       "z\n",
       "\" id=\"DejaVuSans-52\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(92.144054 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-52\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_4\">\n",
       "     <g id=\"line2d_4\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"136.329383\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_4\">\n",
       "      <!-- 0.6 -->\n",
       "      <defs>\n",
       "       <path d=\"M 33.015625 40.375 \n",
       "Q 26.375 40.375 22.484375 35.828125 \n",
       "Q 18.609375 31.296875 18.609375 23.390625 \n",
       "Q 18.609375 15.53125 22.484375 10.953125 \n",
       "Q 26.375 6.390625 33.015625 6.390625 \n",
       "Q 39.65625 6.390625 43.53125 10.953125 \n",
       "Q 47.40625 15.53125 47.40625 23.390625 \n",
       "Q 47.40625 31.296875 43.53125 35.828125 \n",
       "Q 39.65625 40.375 33.015625 40.375 \n",
       "z\n",
       "M 52.59375 71.296875 \n",
       "L 52.59375 62.3125 \n",
       "Q 48.875 64.0625 45.09375 64.984375 \n",
       "Q 41.3125 65.921875 37.59375 65.921875 \n",
       "Q 27.828125 65.921875 22.671875 59.328125 \n",
       "Q 17.53125 52.734375 16.796875 39.40625 \n",
       "Q 19.671875 43.65625 24.015625 45.921875 \n",
       "Q 28.375 48.1875 33.59375 48.1875 \n",
       "Q 44.578125 48.1875 50.953125 41.515625 \n",
       "Q 57.328125 34.859375 57.328125 23.390625 \n",
       "Q 57.328125 12.15625 50.6875 5.359375 \n",
       "Q 44.046875 -1.421875 33.015625 -1.421875 \n",
       "Q 20.359375 -1.421875 13.671875 8.265625 \n",
       "Q 6.984375 17.96875 6.984375 36.375 \n",
       "Q 6.984375 53.65625 15.1875 63.9375 \n",
       "Q 23.390625 74.21875 37.203125 74.21875 \n",
       "Q 40.921875 74.21875 44.703125 73.484375 \n",
       "Q 48.484375 72.75 52.59375 71.296875 \n",
       "z\n",
       "\" id=\"DejaVuSans-54\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(128.377821 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-54\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_5\">\n",
       "     <g id=\"line2d_5\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"172.563149\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_5\">\n",
       "      <!-- 0.8 -->\n",
       "      <defs>\n",
       "       <path d=\"M 31.78125 34.625 \n",
       "Q 24.75 34.625 20.71875 30.859375 \n",
       "Q 16.703125 27.09375 16.703125 20.515625 \n",
       "Q 16.703125 13.921875 20.71875 10.15625 \n",
       "Q 24.75 6.390625 31.78125 6.390625 \n",
       "Q 38.8125 6.390625 42.859375 10.171875 \n",
       "Q 46.921875 13.96875 46.921875 20.515625 \n",
       "Q 46.921875 27.09375 42.890625 30.859375 \n",
       "Q 38.875 34.625 31.78125 34.625 \n",
       "z\n",
       "M 21.921875 38.8125 \n",
       "Q 15.578125 40.375 12.03125 44.71875 \n",
       "Q 8.5 49.078125 8.5 55.328125 \n",
       "Q 8.5 64.0625 14.71875 69.140625 \n",
       "Q 20.953125 74.21875 31.78125 74.21875 \n",
       "Q 42.671875 74.21875 48.875 69.140625 \n",
       "Q 55.078125 64.0625 55.078125 55.328125 \n",
       "Q 55.078125 49.078125 51.53125 44.71875 \n",
       "Q 48 40.375 41.703125 38.8125 \n",
       "Q 48.828125 37.15625 52.796875 32.3125 \n",
       "Q 56.78125 27.484375 56.78125 20.515625 \n",
       "Q 56.78125 9.90625 50.3125 4.234375 \n",
       "Q 43.84375 -1.421875 31.78125 -1.421875 \n",
       "Q 19.734375 -1.421875 13.25 4.234375 \n",
       "Q 6.78125 9.90625 6.78125 20.515625 \n",
       "Q 6.78125 27.484375 10.78125 32.3125 \n",
       "Q 14.796875 37.15625 21.921875 38.8125 \n",
       "z\n",
       "M 18.3125 54.390625 \n",
       "Q 18.3125 48.734375 21.84375 45.5625 \n",
       "Q 25.390625 42.390625 31.78125 42.390625 \n",
       "Q 38.140625 42.390625 41.71875 45.5625 \n",
       "Q 45.3125 48.734375 45.3125 54.390625 \n",
       "Q 45.3125 60.0625 41.71875 63.234375 \n",
       "Q 38.140625 66.40625 31.78125 66.40625 \n",
       "Q 25.390625 66.40625 21.84375 63.234375 \n",
       "Q 18.3125 60.0625 18.3125 54.390625 \n",
       "z\n",
       "\" id=\"DejaVuSans-56\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(164.611587 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-56\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_6\">\n",
       "     <g id=\"line2d_6\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"208.796916\" xlink:href=\"#m805066baff\" y=\"143.1\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_6\">\n",
       "      <!-- 1.0 -->\n",
       "      <defs>\n",
       "       <path d=\"M 12.40625 8.296875 \n",
       "L 28.515625 8.296875 \n",
       "L 28.515625 63.921875 \n",
       "L 10.984375 60.40625 \n",
       "L 10.984375 69.390625 \n",
       "L 28.421875 72.90625 \n",
       "L 38.28125 72.90625 \n",
       "L 38.28125 8.296875 \n",
       "L 54.390625 8.296875 \n",
       "L 54.390625 0 \n",
       "L 12.40625 0 \n",
       "z\n",
       "\" id=\"DejaVuSans-49\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(200.845353 157.698438)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "       <use x=\"63.623047\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "       <use x=\"95.410156\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_2\">\n",
       "    <g id=\"ytick_1\">\n",
       "     <g id=\"line2d_7\">\n",
       "      <defs>\n",
       "       <path d=\"M 0 0 \n",
       "L -3.5 0 \n",
       "\" id=\"m20fc77d18f\" style=\"stroke:#000000;stroke-width:0.8;\"/>\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"20.5625\" xlink:href=\"#m20fc77d18f\" y=\"136.922727\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_7\">\n",
       "      <!-- 0 -->\n",
       "      <g transform=\"translate(7.2 140.721946)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-48\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_2\">\n",
       "     <g id=\"line2d_8\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"20.5625\" xlink:href=\"#m20fc77d18f\" y=\"110.095173\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_8\">\n",
       "      <!-- 1 -->\n",
       "      <g transform=\"translate(7.2 113.894391)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-49\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_3\">\n",
       "     <g id=\"line2d_9\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"20.5625\" xlink:href=\"#m20fc77d18f\" y=\"83.267618\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_9\">\n",
       "      <!-- 2 -->\n",
       "      <g transform=\"translate(7.2 87.066837)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-50\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_4\">\n",
       "     <g id=\"line2d_10\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"20.5625\" xlink:href=\"#m20fc77d18f\" y=\"56.440064\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_10\">\n",
       "      <!-- 3 -->\n",
       "      <defs>\n",
       "       <path d=\"M 40.578125 39.3125 \n",
       "Q 47.65625 37.796875 51.625 33 \n",
       "Q 55.609375 28.21875 55.609375 21.1875 \n",
       "Q 55.609375 10.40625 48.1875 4.484375 \n",
       "Q 40.765625 -1.421875 27.09375 -1.421875 \n",
       "Q 22.515625 -1.421875 17.65625 -0.515625 \n",
       "Q 12.796875 0.390625 7.625 2.203125 \n",
       "L 7.625 11.71875 \n",
       "Q 11.71875 9.328125 16.59375 8.109375 \n",
       "Q 21.484375 6.890625 26.8125 6.890625 \n",
       "Q 36.078125 6.890625 40.9375 10.546875 \n",
       "Q 45.796875 14.203125 45.796875 21.1875 \n",
       "Q 45.796875 27.640625 41.28125 31.265625 \n",
       "Q 36.765625 34.90625 28.71875 34.90625 \n",
       "L 20.21875 34.90625 \n",
       "L 20.21875 43.015625 \n",
       "L 29.109375 43.015625 \n",
       "Q 36.375 43.015625 40.234375 45.921875 \n",
       "Q 44.09375 48.828125 44.09375 54.296875 \n",
       "Q 44.09375 59.90625 40.109375 62.90625 \n",
       "Q 36.140625 65.921875 28.71875 65.921875 \n",
       "Q 24.65625 65.921875 20.015625 65.03125 \n",
       "Q 15.375 64.15625 9.8125 62.3125 \n",
       "L 9.8125 71.09375 \n",
       "Q 15.4375 72.65625 20.34375 73.4375 \n",
       "Q 25.25 74.21875 29.59375 74.21875 \n",
       "Q 40.828125 74.21875 47.359375 69.109375 \n",
       "Q 53.90625 64.015625 53.90625 55.328125 \n",
       "Q 53.90625 49.265625 50.4375 45.09375 \n",
       "Q 46.96875 40.921875 40.578125 39.3125 \n",
       "z\n",
       "\" id=\"DejaVuSans-51\"/>\n",
       "      </defs>\n",
       "      <g transform=\"translate(7.2 60.239282)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-51\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_5\">\n",
       "     <g id=\"line2d_11\">\n",
       "      <g>\n",
       "       <use style=\"stroke:#000000;stroke-width:0.8;\" x=\"20.5625\" xlink:href=\"#m20fc77d18f\" y=\"29.612509\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_11\">\n",
       "      <!-- 4 -->\n",
       "      <g transform=\"translate(7.2 33.411728)scale(0.1 -0.1)\">\n",
       "       <use xlink:href=\"#DejaVuSans-52\"/>\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"line2d_12\">\n",
       "    <path clip-path=\"url(#p23e4e4cf58)\" d=\"M 29.439773 13.377273 \n",
       "L 31.251461 31.972717 \n",
       "L 33.063149 42.850354 \n",
       "L 34.874838 50.56816 \n",
       "L 36.686526 56.554556 \n",
       "L 38.498214 61.445798 \n",
       "L 40.309903 65.581283 \n",
       "L 42.121591 69.163604 \n",
       "L 43.933279 72.323435 \n",
       "L 45.744968 75.15 \n",
       "L 47.556656 77.706939 \n",
       "L 49.368344 80.041242 \n",
       "L 51.180032 82.188592 \n",
       "L 52.991721 84.176727 \n",
       "L 54.803409 86.027637 \n",
       "L 56.615097 87.759048 \n",
       "L 58.426786 89.385458 \n",
       "L 60.238474 90.918879 \n",
       "L 62.050162 92.36937 \n",
       "L 63.861851 93.745444 \n",
       "L 65.673539 95.054365 \n",
       "L 67.485227 96.302383 \n",
       "L 69.296916 97.494915 \n",
       "L 71.108604 98.636685 \n",
       "L 72.920292 99.73184 \n",
       "L 74.731981 100.784035 \n",
       "L 76.543669 101.796516 \n",
       "L 78.355357 102.772171 \n",
       "L 80.167045 103.713585 \n",
       "L 81.978734 104.623081 \n",
       "L 83.790422 105.502752 \n",
       "L 85.60211 106.354492 \n",
       "L 87.413799 107.18002 \n",
       "L 89.225487 107.980902 \n",
       "L 91.037175 108.758567 \n",
       "L 92.848864 109.514323 \n",
       "L 94.660552 110.24937 \n",
       "L 96.47224 110.964814 \n",
       "L 98.283929 111.661673 \n",
       "L 100.095617 112.340888 \n",
       "L 101.907305 113.00333 \n",
       "L 103.718994 113.649808 \n",
       "L 105.530682 114.281074 \n",
       "L 107.34237 114.897827 \n",
       "L 109.154058 115.500718 \n",
       "L 110.965747 116.090359 \n",
       "L 112.777435 116.667318 \n",
       "L 114.589123 117.232129 \n",
       "L 116.400812 117.785294 \n",
       "L 118.2125 118.327283 \n",
       "L 120.024188 118.85854 \n",
       "L 121.835877 119.379479 \n",
       "L 123.647565 119.890496 \n",
       "L 125.459253 120.39196 \n",
       "L 127.270942 120.884223 \n",
       "L 129.08263 121.367615 \n",
       "L 130.894318 121.842451 \n",
       "L 132.706006 122.309029 \n",
       "L 134.517695 122.767631 \n",
       "L 136.329383 123.218525 \n",
       "L 138.141071 123.661966 \n",
       "L 139.95276 124.098196 \n",
       "L 141.764448 124.527446 \n",
       "L 143.576136 124.949936 \n",
       "L 145.387825 125.365875 \n",
       "L 147.199513 125.775464 \n",
       "L 149.011201 126.178893 \n",
       "L 150.82289 126.576346 \n",
       "L 152.634578 126.967996 \n",
       "L 154.446266 127.354011 \n",
       "L 156.257955 127.73455 \n",
       "L 158.069643 128.109766 \n",
       "L 159.881331 128.479808 \n",
       "L 161.693019 128.844814 \n",
       "L 163.504708 129.204921 \n",
       "L 165.316396 129.560258 \n",
       "L 167.128084 129.91095 \n",
       "L 168.939773 130.257117 \n",
       "L 170.751461 130.598874 \n",
       "L 172.563149 130.936331 \n",
       "L 174.374838 131.269597 \n",
       "L 176.186526 131.598774 \n",
       "L 177.998214 131.92396 \n",
       "L 179.809903 132.245252 \n",
       "L 181.621591 132.562742 \n",
       "L 183.433279 132.876518 \n",
       "L 185.244968 133.186667 \n",
       "L 187.056656 133.493271 \n",
       "L 188.868344 133.79641 \n",
       "L 190.680032 134.096162 \n",
       "L 192.491721 134.392602 \n",
       "L 194.303409 134.685803 \n",
       "L 196.115097 134.975833 \n",
       "L 197.926786 135.262762 \n",
       "L 199.738474 135.546654 \n",
       "L 201.550162 135.827573 \n",
       "L 203.361851 136.105581 \n",
       "L 205.173539 136.380738 \n",
       "L 206.985227 136.653101 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_13\">\n",
       "    <path clip-path=\"url(#p23e4e4cf58)\" d=\"M 29.439773 14.612727 \n",
       "L 31.251461 34.071717 \n",
       "L 33.063149 45.672525 \n",
       "L 34.874838 54.022343 \n",
       "L 36.686526 60.572965 \n",
       "L 38.498214 65.974413 \n",
       "L 40.309903 70.575185 \n",
       "L 42.121591 74.584334 \n",
       "L 43.933279 78.137371 \n",
       "L 45.744968 81.327273 \n",
       "L 47.556656 84.220676 \n",
       "L 49.368344 86.86702 \n",
       "L 51.180032 89.304029 \n",
       "L 52.991721 91.561167 \n",
       "L 54.803409 93.661901 \n",
       "L 56.615097 95.625237 \n",
       "L 58.426786 97.466794 \n",
       "L 60.238474 99.199572 \n",
       "L 62.050162 100.834508 \n",
       "L 63.861851 102.380901 \n",
       "L 65.673539 103.846721 \n",
       "L 67.485227 105.238859 \n",
       "L 69.296916 106.563312 \n",
       "L 71.108604 107.825335 \n",
       "L 72.920292 109.029562 \n",
       "L 74.731981 110.180095 \n",
       "L 76.543669 111.280593 \n",
       "L 78.355357 112.334327 \n",
       "L 80.167045 113.344237 \n",
       "L 81.978734 114.312975 \n",
       "L 83.790422 115.242944 \n",
       "L 85.60211 116.136327 \n",
       "L 87.413799 116.995114 \n",
       "L 89.225487 117.821123 \n",
       "L 91.037175 118.616023 \n",
       "L 92.848864 119.381348 \n",
       "L 94.660552 120.118512 \n",
       "L 96.47224 120.828821 \n",
       "L 98.283929 121.513484 \n",
       "L 100.095617 122.173623 \n",
       "L 101.907305 122.810283 \n",
       "L 103.718994 123.424434 \n",
       "L 105.530682 124.016985 \n",
       "L 107.34237 124.588783 \n",
       "L 109.154058 125.140622 \n",
       "L 110.965747 125.673248 \n",
       "L 112.777435 126.18736 \n",
       "L 114.589123 126.683616 \n",
       "L 116.400812 127.162636 \n",
       "L 118.2125 127.625005 \n",
       "L 120.024188 128.071275 \n",
       "L 121.835877 128.501968 \n",
       "L 123.647565 128.917578 \n",
       "L 125.459253 129.318574 \n",
       "L 127.270942 129.7054 \n",
       "L 129.08263 130.078478 \n",
       "L 130.894318 130.438209 \n",
       "L 132.706006 130.784974 \n",
       "L 134.517695 131.119138 \n",
       "L 136.329383 131.441046 \n",
       "L 138.141071 131.75103 \n",
       "L 139.95276 132.049405 \n",
       "L 141.764448 132.336473 \n",
       "L 143.576136 132.612522 \n",
       "L 145.387825 132.877829 \n",
       "L 147.199513 133.132658 \n",
       "L 149.011201 133.377262 \n",
       "L 150.82289 133.611885 \n",
       "L 152.634578 133.836761 \n",
       "L 154.446266 134.052112 \n",
       "L 156.257955 134.258156 \n",
       "L 158.069643 134.455098 \n",
       "L 159.881331 134.643139 \n",
       "L 161.693019 134.82247 \n",
       "L 163.504708 134.993276 \n",
       "L 165.316396 135.155735 \n",
       "L 167.128084 135.310018 \n",
       "L 168.939773 135.456293 \n",
       "L 170.751461 135.594718 \n",
       "L 172.563149 135.725448 \n",
       "L 174.374838 135.848633 \n",
       "L 176.186526 135.964416 \n",
       "L 177.998214 136.072937 \n",
       "L 179.809903 136.174331 \n",
       "L 181.621591 136.268729 \n",
       "L 183.433279 136.356258 \n",
       "L 185.244968 136.437039 \n",
       "L 187.056656 136.511192 \n",
       "L 188.868344 136.578832 \n",
       "L 190.680032 136.640071 \n",
       "L 192.491721 136.695016 \n",
       "L 194.303409 136.743773 \n",
       "L 196.115097 136.786445 \n",
       "L 197.926786 136.823129 \n",
       "L 199.738474 136.853924 \n",
       "L 201.550162 136.878921 \n",
       "L 203.361851 136.898213 \n",
       "L 205.173539 136.911887 \n",
       "L 206.985227 136.920031 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-dasharray:5.55,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"line2d_14\">\n",
       "    <path clip-path=\"url(#p23e4e4cf58)\" d=\"M 29.439773 19.432229 \n",
       "L 31.251461 42.05623 \n",
       "L 33.063149 56.139579 \n",
       "L 34.874838 66.511571 \n",
       "L 36.686526 74.735368 \n",
       "L 38.498214 81.529905 \n",
       "L 40.309903 87.291315 \n",
       "L 42.121591 92.263941 \n",
       "L 43.933279 96.610692 \n",
       "L 45.744968 100.44655 \n",
       "L 47.556656 103.856279 \n",
       "L 49.368344 106.904552 \n",
       "L 51.180032 109.642089 \n",
       "L 52.991721 112.109584 \n",
       "L 54.803409 114.340305 \n",
       "L 56.615097 116.361889 \n",
       "L 58.426786 118.197604 \n",
       "L 60.238474 119.867268 \n",
       "L 62.050162 121.387932 \n",
       "L 63.861851 122.774395 \n",
       "L 65.673539 124.039596 \n",
       "L 67.485227 125.194926 \n",
       "L 69.296916 126.25047 \n",
       "L 71.108604 127.215204 \n",
       "L 72.920292 128.097155 \n",
       "L 74.731981 128.903528 \n",
       "L 76.543669 129.640812 \n",
       "L 78.355357 130.314876 \n",
       "L 80.167045 130.931036 \n",
       "L 81.978734 131.494126 \n",
       "L 83.790422 132.008545 \n",
       "L 85.60211 132.478309 \n",
       "L 87.413799 132.90709 \n",
       "L 89.225487 133.298248 \n",
       "L 91.037175 133.654866 \n",
       "L 92.848864 133.979772 \n",
       "L 94.660552 134.275566 \n",
       "L 96.47224 134.544638 \n",
       "L 98.283929 134.789188 \n",
       "L 100.095617 135.011243 \n",
       "L 101.907305 135.212671 \n",
       "L 103.718994 135.395193 \n",
       "L 105.530682 135.560397 \n",
       "L 107.34237 135.709746 \n",
       "L 109.154058 135.844591 \n",
       "L 110.965747 135.966178 \n",
       "L 112.777435 136.075655 \n",
       "L 114.589123 136.174083 \n",
       "L 116.400812 136.262438 \n",
       "L 118.2125 136.34162 \n",
       "L 120.024188 136.412459 \n",
       "L 121.835877 136.475718 \n",
       "L 123.647565 136.532102 \n",
       "L 125.459253 136.582255 \n",
       "L 127.270942 136.626772 \n",
       "L 129.08263 136.666198 \n",
       "L 130.894318 136.701034 \n",
       "L 132.706006 136.731739 \n",
       "L 134.517695 136.758732 \n",
       "L 136.329383 136.782396 \n",
       "L 138.141071 136.803083 \n",
       "L 139.95276 136.821112 \n",
       "L 141.764448 136.836773 \n",
       "L 143.576136 136.850332 \n",
       "L 145.387825 136.862029 \n",
       "L 147.199513 136.872079 \n",
       "L 149.011201 136.880681 \n",
       "L 150.82289 136.888011 \n",
       "L 152.634578 136.894228 \n",
       "L 154.446266 136.899475 \n",
       "L 156.257955 136.903881 \n",
       "L 158.069643 136.90756 \n",
       "L 159.881331 136.910613 \n",
       "L 161.693019 136.91313 \n",
       "L 163.504708 136.91519 \n",
       "L 165.316396 136.916865 \n",
       "L 167.128084 136.918214 \n",
       "L 168.939773 136.919292 \n",
       "L 170.751461 136.920145 \n",
       "L 172.563149 136.920812 \n",
       "L 174.374838 136.921328 \n",
       "L 176.186526 136.921721 \n",
       "L 177.998214 136.922018 \n",
       "L 179.809903 136.922237 \n",
       "L 181.621591 136.922396 \n",
       "L 183.433279 136.92251 \n",
       "L 185.244968 136.922589 \n",
       "L 187.056656 136.922642 \n",
       "L 188.868344 136.922677 \n",
       "L 190.680032 136.922699 \n",
       "L 192.491721 136.922712 \n",
       "L 194.303409 136.92272 \n",
       "L 196.115097 136.922724 \n",
       "L 197.926786 136.922726 \n",
       "L 199.738474 136.922727 \n",
       "L 201.550162 136.922727 \n",
       "L 203.361851 136.922727 \n",
       "L 205.173539 136.922727 \n",
       "L 206.985227 136.922727 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-dasharray:9.6,2.4,1.5,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 20.5625 143.1 \n",
       "L 20.5625 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_4\">\n",
       "    <path d=\"M 215.8625 143.1 \n",
       "L 215.8625 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_5\">\n",
       "    <path d=\"M 20.5625 143.1 \n",
       "L 215.8625 143.1 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"patch_6\">\n",
       "    <path d=\"M 20.5625 7.2 \n",
       "L 215.8625 7.2 \n",
       "\" style=\"fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;\"/>\n",
       "   </g>\n",
       "   <g id=\"legend_1\">\n",
       "    <g id=\"patch_7\">\n",
       "     <path d=\"M 114.49375 59.234375 \n",
       "L 208.8625 59.234375 \n",
       "Q 210.8625 59.234375 210.8625 57.234375 \n",
       "L 210.8625 14.2 \n",
       "Q 210.8625 12.2 208.8625 12.2 \n",
       "L 114.49375 12.2 \n",
       "Q 112.49375 12.2 112.49375 14.2 \n",
       "L 112.49375 57.234375 \n",
       "Q 112.49375 59.234375 114.49375 59.234375 \n",
       "z\n",
       "\" style=\"fill:#ffffff;opacity:0.8;stroke:#cccccc;stroke-linejoin:miter;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_15\">\n",
       "     <path d=\"M 116.49375 20.298438 \n",
       "L 136.49375 20.298438 \n",
       "\" style=\"fill:none;stroke:#1f77b4;stroke-linecap:square;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_16\"/>\n",
       "    <g id=\"text_12\">\n",
       "     <!-- gamma=0.0 -->\n",
       "     <defs>\n",
       "      <path d=\"M 45.40625 27.984375 \n",
       "Q 45.40625 37.75 41.375 43.109375 \n",
       "Q 37.359375 48.484375 30.078125 48.484375 \n",
       "Q 22.859375 48.484375 18.828125 43.109375 \n",
       "Q 14.796875 37.75 14.796875 27.984375 \n",
       "Q 14.796875 18.265625 18.828125 12.890625 \n",
       "Q 22.859375 7.515625 30.078125 7.515625 \n",
       "Q 37.359375 7.515625 41.375 12.890625 \n",
       "Q 45.40625 18.265625 45.40625 27.984375 \n",
       "z\n",
       "M 54.390625 6.78125 \n",
       "Q 54.390625 -7.171875 48.1875 -13.984375 \n",
       "Q 42 -20.796875 29.203125 -20.796875 \n",
       "Q 24.46875 -20.796875 20.265625 -20.09375 \n",
       "Q 16.0625 -19.390625 12.109375 -17.921875 \n",
       "L 12.109375 -9.1875 \n",
       "Q 16.0625 -11.328125 19.921875 -12.34375 \n",
       "Q 23.78125 -13.375 27.78125 -13.375 \n",
       "Q 36.625 -13.375 41.015625 -8.765625 \n",
       "Q 45.40625 -4.15625 45.40625 5.171875 \n",
       "L 45.40625 9.625 \n",
       "Q 42.625 4.78125 38.28125 2.390625 \n",
       "Q 33.9375 0 27.875 0 \n",
       "Q 17.828125 0 11.671875 7.65625 \n",
       "Q 5.515625 15.328125 5.515625 27.984375 \n",
       "Q 5.515625 40.671875 11.671875 48.328125 \n",
       "Q 17.828125 56 27.875 56 \n",
       "Q 33.9375 56 38.28125 53.609375 \n",
       "Q 42.625 51.21875 45.40625 46.390625 \n",
       "L 45.40625 54.6875 \n",
       "L 54.390625 54.6875 \n",
       "z\n",
       "\" id=\"DejaVuSans-103\"/>\n",
       "      <path d=\"M 34.28125 27.484375 \n",
       "Q 23.390625 27.484375 19.1875 25 \n",
       "Q 14.984375 22.515625 14.984375 16.5 \n",
       "Q 14.984375 11.71875 18.140625 8.90625 \n",
       "Q 21.296875 6.109375 26.703125 6.109375 \n",
       "Q 34.1875 6.109375 38.703125 11.40625 \n",
       "Q 43.21875 16.703125 43.21875 25.484375 \n",
       "L 43.21875 27.484375 \n",
       "z\n",
       "M 52.203125 31.203125 \n",
       "L 52.203125 0 \n",
       "L 43.21875 0 \n",
       "L 43.21875 8.296875 \n",
       "Q 40.140625 3.328125 35.546875 0.953125 \n",
       "Q 30.953125 -1.421875 24.3125 -1.421875 \n",
       "Q 15.921875 -1.421875 10.953125 3.296875 \n",
       "Q 6 8.015625 6 15.921875 \n",
       "Q 6 25.140625 12.171875 29.828125 \n",
       "Q 18.359375 34.515625 30.609375 34.515625 \n",
       "L 43.21875 34.515625 \n",
       "L 43.21875 35.40625 \n",
       "Q 43.21875 41.609375 39.140625 45 \n",
       "Q 35.0625 48.390625 27.6875 48.390625 \n",
       "Q 23 48.390625 18.546875 47.265625 \n",
       "Q 14.109375 46.140625 10.015625 43.890625 \n",
       "L 10.015625 52.203125 \n",
       "Q 14.9375 54.109375 19.578125 55.046875 \n",
       "Q 24.21875 56 28.609375 56 \n",
       "Q 40.484375 56 46.34375 49.84375 \n",
       "Q 52.203125 43.703125 52.203125 31.203125 \n",
       "z\n",
       "\" id=\"DejaVuSans-97\"/>\n",
       "      <path d=\"M 52 44.1875 \n",
       "Q 55.375 50.25 60.0625 53.125 \n",
       "Q 64.75 56 71.09375 56 \n",
       "Q 79.640625 56 84.28125 50.015625 \n",
       "Q 88.921875 44.046875 88.921875 33.015625 \n",
       "L 88.921875 0 \n",
       "L 79.890625 0 \n",
       "L 79.890625 32.71875 \n",
       "Q 79.890625 40.578125 77.09375 44.375 \n",
       "Q 74.3125 48.1875 68.609375 48.1875 \n",
       "Q 61.625 48.1875 57.5625 43.546875 \n",
       "Q 53.515625 38.921875 53.515625 30.90625 \n",
       "L 53.515625 0 \n",
       "L 44.484375 0 \n",
       "L 44.484375 32.71875 \n",
       "Q 44.484375 40.625 41.703125 44.40625 \n",
       "Q 38.921875 48.1875 33.109375 48.1875 \n",
       "Q 26.21875 48.1875 22.15625 43.53125 \n",
       "Q 18.109375 38.875 18.109375 30.90625 \n",
       "L 18.109375 0 \n",
       "L 9.078125 0 \n",
       "L 9.078125 54.6875 \n",
       "L 18.109375 54.6875 \n",
       "L 18.109375 46.1875 \n",
       "Q 21.1875 51.21875 25.484375 53.609375 \n",
       "Q 29.78125 56 35.6875 56 \n",
       "Q 41.65625 56 45.828125 52.96875 \n",
       "Q 50 49.953125 52 44.1875 \n",
       "z\n",
       "\" id=\"DejaVuSans-109\"/>\n",
       "      <path d=\"M 10.59375 45.40625 \n",
       "L 73.1875 45.40625 \n",
       "L 73.1875 37.203125 \n",
       "L 10.59375 37.203125 \n",
       "z\n",
       "M 10.59375 25.484375 \n",
       "L 73.1875 25.484375 \n",
       "L 73.1875 17.1875 \n",
       "L 10.59375 17.1875 \n",
       "z\n",
       "\" id=\"DejaVuSans-61\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(144.49375 23.798438)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"63.476562\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"124.755859\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"222.167969\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"319.580078\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"380.859375\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"464.648438\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "      <use x=\"528.271484\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"560.058594\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_17\">\n",
       "     <path d=\"M 116.49375 34.976562 \n",
       "L 136.49375 34.976562 \n",
       "\" style=\"fill:none;stroke:#ff7f0e;stroke-dasharray:5.55,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_18\"/>\n",
       "    <g id=\"text_13\">\n",
       "     <!-- gamma=1.0 -->\n",
       "     <g transform=\"translate(144.49375 38.476562)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"63.476562\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"124.755859\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"222.167969\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"319.580078\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"380.859375\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"464.648438\" xlink:href=\"#DejaVuSans-49\"/>\n",
       "      <use x=\"528.271484\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"560.058594\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"line2d_19\">\n",
       "     <path d=\"M 116.49375 49.654688 \n",
       "L 136.49375 49.654688 \n",
       "\" style=\"fill:none;stroke:#2ca02c;stroke-dasharray:9.6,2.4,1.5,2.4;stroke-dashoffset:0;stroke-width:1.5;\"/>\n",
       "    </g>\n",
       "    <g id=\"line2d_20\"/>\n",
       "    <g id=\"text_14\">\n",
       "     <!-- gamma=5.0 -->\n",
       "     <defs>\n",
       "      <path d=\"M 10.796875 72.90625 \n",
       "L 49.515625 72.90625 \n",
       "L 49.515625 64.59375 \n",
       "L 19.828125 64.59375 \n",
       "L 19.828125 46.734375 \n",
       "Q 21.96875 47.46875 24.109375 47.828125 \n",
       "Q 26.265625 48.1875 28.421875 48.1875 \n",
       "Q 40.625 48.1875 47.75 41.5 \n",
       "Q 54.890625 34.8125 54.890625 23.390625 \n",
       "Q 54.890625 11.625 47.5625 5.09375 \n",
       "Q 40.234375 -1.421875 26.90625 -1.421875 \n",
       "Q 22.3125 -1.421875 17.546875 -0.640625 \n",
       "Q 12.796875 0.140625 7.71875 1.703125 \n",
       "L 7.71875 11.625 \n",
       "Q 12.109375 9.234375 16.796875 8.0625 \n",
       "Q 21.484375 6.890625 26.703125 6.890625 \n",
       "Q 35.15625 6.890625 40.078125 11.328125 \n",
       "Q 45.015625 15.765625 45.015625 23.390625 \n",
       "Q 45.015625 31 40.078125 35.4375 \n",
       "Q 35.15625 39.890625 26.703125 39.890625 \n",
       "Q 22.75 39.890625 18.8125 39.015625 \n",
       "Q 14.890625 38.140625 10.796875 36.28125 \n",
       "z\n",
       "\" id=\"DejaVuSans-53\"/>\n",
       "     </defs>\n",
       "     <g transform=\"translate(144.49375 53.154688)scale(0.1 -0.1)\">\n",
       "      <use xlink:href=\"#DejaVuSans-103\"/>\n",
       "      <use x=\"63.476562\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"124.755859\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"222.167969\" xlink:href=\"#DejaVuSans-109\"/>\n",
       "      <use x=\"319.580078\" xlink:href=\"#DejaVuSans-97\"/>\n",
       "      <use x=\"380.859375\" xlink:href=\"#DejaVuSans-61\"/>\n",
       "      <use x=\"464.648438\" xlink:href=\"#DejaVuSans-53\"/>\n",
       "      <use x=\"528.271484\" xlink:href=\"#DejaVuSans-46\"/>\n",
       "      <use x=\"560.058594\" xlink:href=\"#DejaVuSans-48\"/>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"p23e4e4cf58\">\n",
       "   <rect height=\"135.9\" width=\"195.3\" x=\"20.5625\" y=\"7.2\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<Figure size 252x180 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def focal_loss(gamma, x):\n",
    "    return -(1 - x) ** gamma * np.log(x)\n",
    "\n",
    "x = np.arange(0.01, 1, 0.01)\n",
    "\n",
    "for l, gamma in zip(lines, [0, 1, 5]):\n",
    "    y = d2l.plt.plot(x, focal_loss(gamma, x), l,\n",
    "                     label='gamma=%.1f' % gamma)\n",
    "    \n",
    "d2l.plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and Prediction\n",
    "\n",
    "* When an object is relatively large compared to the image, the model normally adopts a larger input image size.\n",
    "* This generally produces a large number of negative anchor boxes when labeling anchor box categories. We can sample the negative anchor boxes to better balance the data categories.\n",
    "* Assign hyper-parameters with different weights to the anchor box category loss and positive anchor box offset loss in the loss function.\n",
    "* Refer to the SSD paper. What methods can be used to evaluate the precision of\n",
    "  object detection models :cite:`Liu.Anguelov.Erhan.ea.2016`?"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}