4. Train-1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.preprocessing import image\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import pandas as pd \n",
    "import numpy as np \n",
    "from tqdm import tqdm\n",
    "import matplotlib.pyplot as plt\n",
    "import cv2\n",
    "from keras import models\n",
    "from keras import layers\n",
    "from keras import optimizers\n",
    "from PIL import Image\n",
    "import sklearn\n",
    "from sklearn.model_selection import train_test_split\n",
    "from statistics import mean\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def arrange_data(df):\n",
    "    \n",
    "    image_data = []\n",
    "    img_paths = np.asarray(df.iloc[:, 0]) #First column is the image paths\n",
    "    \n",
    "    for i in tqdm(range(len(img_paths))):\n",
    "              \n",
    "        img = image.load_img(img_paths[i],target_size=(200,150,3))\n",
    "        img = image.img_to_array(img)\n",
    "        img = img/255\n",
    "        image_data.append(img)\n",
    "        \n",
    "        \n",
    "    \n",
    "    X = np.array(image_data)\n",
    "    Y = np.array(df.iloc[:,1:29])\n",
    "    \n",
    "    print(\"Shape of images:\", X.shape)\n",
    "    print(\"Shape of labels:\", Y.shape)\n",
    "    \n",
    "    return X, Y\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 65/25841 [00:00<00:39, 644.50it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing train..\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 25841/25841 [01:50<00:00, 234.47it/s]\n",
      "  1%|          | 78/9966 [00:00<00:12, 778.51it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of images: (25841, 200, 150, 3)\n",
      "Shape of labels: (25841, 28)\n",
      "Processing valid..\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9966/9966 [01:01<00:00, 162.96it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of images: (9966, 200, 150, 3)\n",
      "Shape of labels: (9966, 28)\n"
     ]
    }
   ],
   "source": [
    "print(\"Processing train..\")\n",
    "train_df = pd.read_csv(\"Train.csv\", delimiter=\" \")\n",
    "X_train, Y_train = arrange_data (train_df)\n",
    "\n",
    "print(\"Processing valid..\")\n",
    "val_df = pd.read_csv(\"Valid.csv\", delimiter=\" \")\n",
    "X_val, Y_val = arrange_data (val_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_5 (Conv2D)            (None, 196, 146, 16)      1216      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 98, 73, 16)        0         \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 98, 73, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 94, 69, 32)        12832     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_6 (MaxPooling2 (None, 47, 34, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_8 (Dropout)          (None, 47, 34, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 43, 30, 64)        51264     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2 (None, 21, 15, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_9 (Dropout)          (None, 21, 15, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 17, 11, 64)        102464    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2 (None, 8, 5, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_10 (Dropout)         (None, 8, 5, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 2560)              0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 128)               327808    \n",
      "_________________________________________________________________\n",
      "dropout_11 (Dropout)         (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dropout_12 (Dropout)         (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 28)                1820      \n",
      "=================================================================\n",
      "Total params: 505,660\n",
      "Trainable params: 505,660\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "num_classes = 28  \n",
    "\n",
    "model = Sequential()\n",
    "model.add(Conv2D(filters=16, kernel_size=(5, 5), activation=\"relu\", input_shape=(200,150,3)))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Conv2D(filters=32, kernel_size=(5, 5), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Conv2D(filters=64, kernel_size=(5, 5), activation=\"relu\"))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Conv2D(filters=64, kernel_size=(5, 5), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(64, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes, activation='sigmoid'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:3376: The name tf.log is deprecated. Please use tf.math.log instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/tensorflow_core/python/ops/nn_impl.py:183: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:986: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:973: The name tf.assign is deprecated. Please use tf.compat.v1.assign instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:2741: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.\n",
      "\n",
      "Train on 25841 samples, validate on 9966 samples\n",
      "Epoch 1/30\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:174: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:181: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:190: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:199: The name tf.is_variable_initialized is deprecated. Please use tf.compat.v1.is_variable_initialized instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/dmisra/keras_tf/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py:206: The name tf.variables_initializer is deprecated. Please use tf.compat.v1.variables_initializer instead.\n",
      "\n",
      "25841/25841 [==============================] - 26s 1ms/step - loss: 0.2740 - acc: 0.9018 - val_loss: 0.2145 - val_acc: 0.9234\n",
      "Epoch 2/30\n",
      "25841/25841 [==============================] - 21s 822us/step - loss: 0.2254 - acc: 0.9227 - val_loss: 0.2137 - val_acc: 0.9234\n",
      "Epoch 3/30\n",
      "25841/25841 [==============================] - 22s 844us/step - loss: 0.2213 - acc: 0.9230 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 4/30\n",
      "25841/25841 [==============================] - 22s 847us/step - loss: 0.2197 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 5/30\n",
      "25841/25841 [==============================] - 21s 832us/step - loss: 0.2184 - acc: 0.9230 - val_loss: 0.2135 - val_acc: 0.9234\n",
      "Epoch 6/30\n",
      "25841/25841 [==============================] - 21s 830us/step - loss: 0.2178 - acc: 0.9228 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 7/30\n",
      "25841/25841 [==============================] - 22s 834us/step - loss: 0.2174 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 8/30\n",
      "25841/25841 [==============================] - 21s 825us/step - loss: 0.2170 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 9/30\n",
      "25841/25841 [==============================] - 21s 810us/step - loss: 0.2168 - acc: 0.9229 - val_loss: 0.2135 - val_acc: 0.9234\n",
      "Epoch 10/30\n",
      "25841/25841 [==============================] - 22s 840us/step - loss: 0.2166 - acc: 0.9228 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 11/30\n",
      "25841/25841 [==============================] - 22s 841us/step - loss: 0.2165 - acc: 0.9230 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 12/30\n",
      "25841/25841 [==============================] - 21s 832us/step - loss: 0.2164 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 13/30\n",
      "25841/25841 [==============================] - 22s 843us/step - loss: 0.2162 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 14/30\n",
      "25841/25841 [==============================] - 22s 850us/step - loss: 0.2161 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 15/30\n",
      "25841/25841 [==============================] - 22s 837us/step - loss: 0.2162 - acc: 0.9231 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 16/30\n",
      "25841/25841 [==============================] - 22s 838us/step - loss: 0.2159 - acc: 0.9230 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 17/30\n",
      "25841/25841 [==============================] - 22s 850us/step - loss: 0.2159 - acc: 0.9229 - val_loss: 0.2135 - val_acc: 0.9234\n",
      "Epoch 18/30\n",
      "25841/25841 [==============================] - 21s 823us/step - loss: 0.2158 - acc: 0.9228 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 19/30\n",
      "25841/25841 [==============================] - 21s 825us/step - loss: 0.2158 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 20/30\n",
      "25841/25841 [==============================] - 21s 829us/step - loss: 0.2156 - acc: 0.9230 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 21/30\n",
      "25841/25841 [==============================] - 22s 836us/step - loss: 0.2156 - acc: 0.9228 - val_loss: 0.2135 - val_acc: 0.9234\n",
      "Epoch 22/30\n",
      "25841/25841 [==============================] - 22s 836us/step - loss: 0.2156 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 23/30\n",
      "25841/25841 [==============================] - 22s 836us/step - loss: 0.2156 - acc: 0.9229 - val_loss: 0.2137 - val_acc: 0.9228\n",
      "Epoch 24/30\n",
      "25841/25841 [==============================] - 21s 812us/step - loss: 0.2155 - acc: 0.9230 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 25/30\n",
      "25841/25841 [==============================] - 22s 837us/step - loss: 0.2155 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 26/30\n",
      "25841/25841 [==============================] - 21s 804us/step - loss: 0.2154 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 27/30\n",
      "25841/25841 [==============================] - 22s 840us/step - loss: 0.2154 - acc: 0.9226 - val_loss: 0.2136 - val_acc: 0.9234\n",
      "Epoch 28/30\n",
      "25841/25841 [==============================] - 22s 833us/step - loss: 0.2153 - acc: 0.9228 - val_loss: 0.2135 - val_acc: 0.9234\n",
      "Epoch 29/30\n",
      "25841/25841 [==============================] - 21s 816us/step - loss: 0.2153 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9228\n",
      "Epoch 30/30\n",
      "25841/25841 [==============================] - 21s 830us/step - loss: 0.2153 - acc: 0.9229 - val_loss: 0.2136 - val_acc: 0.9228\n"
     ]
    }
   ],
   "source": [
    "#model.compile(loss='binary_crossentropy',\n",
    "              #optimizer=keras.optimizers.Adagrad(),\n",
    "              #metrics=['accuracy'])\n",
    "\n",
    "model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "#model.compile(optimizer=optimizers.RMSprop(lr=1e-4), loss='binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "history = model.fit(X_train, Y_train, epochs=30, validation_data=(X_val, Y_val), batch_size=64)\n",
    "model.save('Model_6c.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = history.history['acc']\n",
    "val_acc = history.history['val_acc']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs = range(len(acc))\n",
    "\n",
    "plt.plot(epochs, acc, 'b', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'b', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}