Dog-Breed-Identification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import cv2\n",
    "from tqdm import tqdm\n",
    "import matplotlib.pyplot as plt \n",
    "import sklearn\n",
    "from sklearn.cross_validation import train_test_split\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 id             breed\n",
      "0  000bec180eb18c7604dcecc8fe0dba07       boston_bull\n",
      "1  001513dfcb2ffafc82cccf4d8bbaba97             dingo\n",
      "2  001cdf01b096e06d78e9e5112d419397          pekinese\n",
      "3  00214f311d5d2247d5dfe4fe24b2303d          bluetick\n",
      "4  0021f9ceb3235effd7fcde7f7538ed62  golden_retriever\n",
      "scottish_deerhound      126\n",
      "maltese_dog             117\n",
      "afghan_hound            116\n",
      "entlebucher             115\n",
      "bernese_mountain_dog    114\n",
      "shih-tzu                112\n",
      "great_pyrenees          111\n",
      "pomeranian              111\n",
      "basenji                 110\n",
      "samoyed                 109\n",
      "Name: breed, dtype: int64\n",
      "(120,)\n",
      "[['000bec180eb18c7604dcecc8fe0dba07' 'boston_bull']\n",
      " ['001513dfcb2ffafc82cccf4d8bbaba97' 'dingo']\n",
      " ['001cdf01b096e06d78e9e5112d419397' 'pekinese']\n",
      " ...\n",
      " ['ffe2ca6c940cddfee68fa3cc6c63213f' 'airedale']\n",
      " ['ffe5f6d8e2bff356e9482a80a6e29aac' 'miniature_pinscher']\n",
      " ['fff43b07992508bc822f33d8ffd902ae' 'chesapeake_bay_retriever']]\n"
     ]
    }
   ],
   "source": [
    "lables = pd.read_csv('C:\\\\Users\\\\Ujjawal\\\\Anaconda3\\\\input_files\\\\dog-breed-dataset\\\\file1\\\\labels.csv')\n",
    "breed_count=lables['breed'].value_counts()\n",
    "print(lables.head())\n",
    "print(breed_count.head(10))\n",
    "print(breed_count.shape)\n",
    "print(lables.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   affenpinscher  afghan_hound  african_hunting_dog  airedale  \\\n",
      "0              0             0                    0         0   \n",
      "1              0             0                    0         0   \n",
      "2              0             0                    0         0   \n",
      "3              0             0                    0         0   \n",
      "4              0             0                    0         0   \n",
      "\n",
      "   american_staffordshire_terrier  appenzeller  australian_terrier  basenji  \\\n",
      "0                               0            0                   0        0   \n",
      "1                               0            0                   0        0   \n",
      "2                               0            0                   0        0   \n",
      "3                               0            0                   0        0   \n",
      "4                               0            0                   0        0   \n",
      "\n",
      "   basset  beagle        ...          toy_poodle  toy_terrier  vizsla  \\\n",
      "0       0       0        ...                   0            0       0   \n",
      "1       0       0        ...                   0            0       0   \n",
      "2       0       0        ...                   0            0       0   \n",
      "3       0       0        ...                   0            0       0   \n",
      "4       0       0        ...                   0            0       0   \n",
      "\n",
      "   walker_hound  weimaraner  welsh_springer_spaniel  \\\n",
      "0             0           0                       0   \n",
      "1             0           0                       0   \n",
      "2             0           0                       0   \n",
      "3             0           0                       0   \n",
      "4             0           0                       0   \n",
      "\n",
      "   west_highland_white_terrier  whippet  wire-haired_fox_terrier  \\\n",
      "0                            0        0                        0   \n",
      "1                            0        0                        0   \n",
      "2                            0        0                        0   \n",
      "3                            0        0                        0   \n",
      "4                            0        0                        0   \n",
      "\n",
      "   yorkshire_terrier  \n",
      "0                  0  \n",
      "1                  0  \n",
      "2                  0  \n",
      "3                  0  \n",
      "4                  0  \n",
      "\n",
      "[5 rows x 120 columns]\n",
      "[[0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " ...\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]]\n"
     ]
    }
   ],
   "source": [
    "#One hot Encoding\n",
    "targets = pd.Series(lables['breed'])\n",
    "encoded = pd.get_dummies(targets, sparse = True)\n",
    "encoded_labels = np.asarray(encoded)\n",
    "print(encoded.head())\n",
    "print(encoded_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████| 10222/10222 [02:03<00:00, 82.66it/s]\n"
     ]
    }
   ],
   "source": [
    "img_rows=128\n",
    "img_cols=128\n",
    "no_channels=1\n",
    "\n",
    "feature_x=[]\n",
    "feature_y=[]\n",
    "i=0\n",
    "for f, img in tqdm(lables.values): # f for format ,jpg\n",
    "    train_img = cv2.imread('C:\\\\Users\\\\Ujjawal\\\\Anaconda3\\\\input_files\\\\dog-breed-dataset//train//{}.jpg'.format(f),0)\n",
    "    label = encoded_labels[i]\n",
    "    resized_imgs = cv2.resize(train_img, (img_rows, img_cols)) \n",
    "    feature_x.append(resized_imgs)\n",
    "    feature_y.append(label)\n",
    "    i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10222, 128, 128)\n",
      "(10222, 128, 128, 1)\n",
      "(10222, 120)\n",
      "(8177, 128, 128, 1)\n",
      "(2045, 128, 128, 1)\n"
     ]
    }
   ],
   "source": [
    "x_train_data = np.array(feature_x, np.float32)/255   # /= 255 for normalisation\n",
    "print(x_train_data.shape)     #just for checking purpose\n",
    "x_train_data = np.expand_dims(x_train_data, axis = 3)\n",
    "print (x_train_data.shape)     #just for checking purpose\n",
    "\n",
    "y_train_data = np.array(feature_y, np.uint8)\n",
    "print (y_train_data.shape)    #just for checking purpose\n",
    "\n",
    "x_train, x_val, y_train, y_val = train_test_split(x_train_data, y_train_data, test_size=0.2, random_state=2)\n",
    "print (x_train.shape)      #just for checking purpose\n",
    "print (x_val.shape)        #just for checking purpose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 id  affenpinscher  afghan_hound  \\\n",
      "0  000621fb3cbb32d8935728e48679680e       0.008333      0.008333   \n",
      "1  00102ee9d8eb90812350685311fe5890       0.008333      0.008333   \n",
      "2  0012a730dfa437f5f3613fb75efcd4ce       0.008333      0.008333   \n",
      "3  001510bc8570bbeee98c8d80c8a95ec1       0.008333      0.008333   \n",
      "4  001a5f3114548acdefa3d4da05474c2e       0.008333      0.008333   \n",
      "\n",
      "   african_hunting_dog  airedale  american_staffordshire_terrier  appenzeller  \\\n",
      "0             0.008333  0.008333                        0.008333     0.008333   \n",
      "1             0.008333  0.008333                        0.008333     0.008333   \n",
      "2             0.008333  0.008333                        0.008333     0.008333   \n",
      "3             0.008333  0.008333                        0.008333     0.008333   \n",
      "4             0.008333  0.008333                        0.008333     0.008333   \n",
      "\n",
      "   australian_terrier   basenji    basset        ...          toy_poodle  \\\n",
      "0            0.008333  0.008333  0.008333        ...            0.008333   \n",
      "1            0.008333  0.008333  0.008333        ...            0.008333   \n",
      "2            0.008333  0.008333  0.008333        ...            0.008333   \n",
      "3            0.008333  0.008333  0.008333        ...            0.008333   \n",
      "4            0.008333  0.008333  0.008333        ...            0.008333   \n",
      "\n",
      "   toy_terrier    vizsla  walker_hound  weimaraner  welsh_springer_spaniel  \\\n",
      "0     0.008333  0.008333      0.008333    0.008333                0.008333   \n",
      "1     0.008333  0.008333      0.008333    0.008333                0.008333   \n",
      "2     0.008333  0.008333      0.008333    0.008333                0.008333   \n",
      "3     0.008333  0.008333      0.008333    0.008333                0.008333   \n",
      "4     0.008333  0.008333      0.008333    0.008333                0.008333   \n",
      "\n",
      "   west_highland_white_terrier   whippet  wire-haired_fox_terrier  \\\n",
      "0                     0.008333  0.008333                 0.008333   \n",
      "1                     0.008333  0.008333                 0.008333   \n",
      "2                     0.008333  0.008333                 0.008333   \n",
      "3                     0.008333  0.008333                 0.008333   \n",
      "4                     0.008333  0.008333                 0.008333   \n",
      "\n",
      "   yorkshire_terrier  \n",
      "0           0.008333  \n",
      "1           0.008333  \n",
      "2           0.008333  \n",
      "3           0.008333  \n",
      "4           0.008333  \n",
      "\n",
      "[5 rows x 121 columns]\n",
      "0    000621fb3cbb32d8935728e48679680e\n",
      "1    00102ee9d8eb90812350685311fe5890\n",
      "2    0012a730dfa437f5f3613fb75efcd4ce\n",
      "3    001510bc8570bbeee98c8d80c8a95ec1\n",
      "4    001a5f3114548acdefa3d4da05474c2e\n",
      "Name: id, dtype: object\n"
     ]
    }
   ],
   "source": [
    "submission = pd.read_csv('C:\\\\Users\\\\Ujjawal\\\\Anaconda3\\\\input_files\\\\dog-breed-dataset\\\\file1\\\\sample_submission.csv')\n",
    "print(submission.head(5))\n",
    "test_img = submission['id']\n",
    "print (test_img.head(5)) #just for checking purpose\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████| 10357/10357 [03:22<00:00, 51.07it/s]\n"
     ]
    }
   ],
   "source": [
    "x_test_feature = []\n",
    "\n",
    "i = 0 # initialisation\n",
    "for f in tqdm(test_img.values): # f for format ,jpg\n",
    "    imgs = cv2.imread('C:\\\\Users\\\\Ujjawal\\\\Anaconda3\\\\input_files\\\\dog-breed-dataset\\\\test\\\\{}.jpg'.format(f), 0)\n",
    "    resized_img = cv2.resize(imgs, (img_rows, img_cols)) \n",
    "    x_test_feature.append(resized_img)\n",
    "      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "#importing keras libray\n",
    "from keras.models import Sequential  # initial NN\n",
    "from keras.layers import Dense, Dropout # construct each layer\n",
    "from keras.layers import Convolution2D # swipe across the image by 1\n",
    "from keras.layers import MaxPooling2D # swipe across by pool size\n",
    "from keras.layers import Flatten"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Buiding our CNN model\n",
    "model = Sequential()   #initializing the CNN\n",
    "\n",
    "#Adding first convolution layer\n",
    "model.add(Convolution2D (filters = 64, kernel_size = (3,3),padding = 'Same', \n",
    "                         activation ='relu', input_shape = (img_rows, img_cols, no_channels)))\n",
    "\n",
    "#Pooling process\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "\n",
    "#Adding second convolution layer\n",
    "model.add(Convolution2D (filters = 64, kernel_size = (3,3),padding = 'Same', \n",
    "                         activation ='relu')) \n",
    "\n",
    "#Pooling process\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "\n",
    "#Flattening process\n",
    "model.add(Flatten()) \n",
    "# full connection  \n",
    "model.add(Dense(units = 120, activation = 'relu')) \n",
    "# output layer\n",
    "model.add(Dense(units = 120, activation = 'softmax')) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 128, 128, 64)      640       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 64, 64, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 64, 64, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 65536)             0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 120)               7864440   \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 120)               14520     \n",
      "=================================================================\n",
      "Total params: 7,916,528\n",
      "Trainable params: 7,916,528\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "#compiling our model\n",
    "model.compile(optimizer = 'adam' , loss = \"categorical_crossentropy\", metrics=[\"accuracy\"]) \n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 8177 samples, validate on 2045 samples\n",
      "Epoch 1/2\n",
      " - 390s - loss: 4.7873 - acc: 0.0110 - val_loss: 4.7865 - val_acc: 0.0078\n",
      "Epoch 2/2\n",
      " - 399s - loss: 4.6734 - acc: 0.0236 - val_loss: 4.6600 - val_acc: 0.0220\n"
     ]
    }
   ],
   "source": [
    "#Training our model\n",
    "batch_size = 128 \n",
    "nb_epochs = 2\n",
    "history = model.fit(x_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    epochs=nb_epochs,\n",
    "                    verbose=2, \n",
    "                    validation_data=(x_val, y_val),\n",
    "                    initial_epoch=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2,1)\n",
    "ax[0].plot(history.history['loss'], color='b', label=\"Training loss\")\n",
    "ax[0].plot(history.history['val_loss'], color='r', label=\"validation loss\",axes =ax[0])\n",
    "legend = ax[0].legend(loc='best', shadow=True)\n",
    "\n",
    "ax[1].plot(history.history['acc'], color='b', label=\"Training accuracy\")\n",
    "ax[1].plot(history.history['val_acc'], color='r',label=\"Validation accuracy\")\n",
    "legend = ax[1].legend(loc='best', shadow=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10357, 128, 128)\n",
      "(10357, 128, 128, 1)\n"
     ]
    }
   ],
   "source": [
    "x_test_data = np.array(x_test_feature, np.float32) / 255. # Normalization of test data\n",
    "print (x_test_data.shape)\n",
    "x_test_data = np.expand_dims(x_test_data, axis = 3)\n",
    "print (x_test_data.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Testing our model on testing image data\n",
    "results = model.predict(x_test_data)\n",
    "prediction = pd.DataFrame(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set column names to those generated by the one-hot encoding earlier\n",
    "col_names = encoded.columns.values\n",
    "prediction.columns = col_names\n",
    "\n",
    "# Insert the column id from the sample_submission at the start of the data frame\n",
    "prediction.insert(0, 'id', submission['id'])\n",
    "\n",
    "\n",
    "submission = prediction\n",
    "submission.to_csv('dog_breed_submission.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}