12. CNN - Dog or cat Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using Theano backend.\n",
      "WARNING (theano.tensor.blas): Using NumPy C-API based implementation for BLAS functions.\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Convolution2D\n",
    "from keras.layers import MaxPooling2D\n",
    "from keras.layers import Flatten"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Intialize the model\n",
    "model=Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add Convolution Layer\n",
    "model.add(Convolution2D(32,(3,3),input_shape=(64,64,3),activation=\"relu\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add Pooling Layer\n",
    "model.add(MaxPooling2D(pool_size = (2, 2)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add Flattening Layer\n",
    "model.add(Flatten())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/ipykernel_launcher.py:2: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", units=120, kernel_initializer=\"uniform\")`\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "#Add Hidden Layer\n",
    "model.add(Dense(init=\"uniform\",activation=\"relu\",output_dim=120))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sudharsan/anaconda3/envs/universe/lib/python3.6/site-packages/ipykernel_launcher.py:2: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"sigmoid\", units=1, kernel_initializer=\"uniform\")`\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "#Add Output layer\n",
    "model.add(Dense(init=\"uniform\",activation=\"sigmoid\",output_dim=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Compile the model\n",
    "model.compile(loss=\"binary_crossentropy\",optimizer=\"adam\",metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "train_datagen = ImageDataGenerator(rescale = 1./255,\n",
    "                                   shear_range = 0.2,\n",
    "                                   zoom_range = 0.2,\n",
    "                                   horizontal_flip = True)\n",
    "\n",
    "test_datagen = ImageDataGenerator(rescale = 1./255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 8010 images belonging to 2 classes.\n",
      "Found 2002 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "x_train = train_datagen.flow_from_directory('dataset/training_set',\n",
    "                                                 target_size = (64, 64),\n",
    "                                                 batch_size = 32,\n",
    "                                                     class_mode = 'binary')\n",
    "x_test = test_datagen.flow_from_directory('dataset/test_set',\n",
    "                                            target_size = (64, 64),\n",
    "                                            batch_size = 32,\n",
    "                                            class_mode = 'binary')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'cats': 0, 'dogs': 1}\n"
     ]
    }
   ],
   "source": [
    "print(x_train.class_indices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "250/250 [==============================] - 179s 718ms/step - loss: 0.6715 - accuracy: 0.5830 - val_loss: 0.5945 - val_accuracy: 0.6618\n",
      "Epoch 2/10\n",
      "250/250 [==============================] - 177s 707ms/step - loss: 0.6133 - accuracy: 0.6647 - val_loss: 0.5780 - val_accuracy: 0.6503\n",
      "Epoch 3/10\n",
      "250/250 [==============================] - 155s 619ms/step - loss: 0.5765 - accuracy: 0.6973 - val_loss: 0.5574 - val_accuracy: 0.7278\n",
      "Epoch 4/10\n",
      "250/250 [==============================] - 149s 598ms/step - loss: 0.5642 - accuracy: 0.7069 - val_loss: 0.6422 - val_accuracy: 0.7018\n",
      "Epoch 5/10\n",
      "250/250 [==============================] - 145s 582ms/step - loss: 0.5335 - accuracy: 0.7276 - val_loss: 0.4835 - val_accuracy: 0.7283\n",
      "Epoch 6/10\n",
      "250/250 [==============================] - 144s 577ms/step - loss: 0.5268 - accuracy: 0.7340 - val_loss: 0.5632 - val_accuracy: 0.7433\n",
      "Epoch 7/10\n",
      "250/250 [==============================] - 153s 613ms/step - loss: 0.5099 - accuracy: 0.7486 - val_loss: 0.3702 - val_accuracy: 0.6873\n",
      "Epoch 8/10\n",
      "250/250 [==============================] - 182s 728ms/step - loss: 0.4935 - accuracy: 0.7580 - val_loss: 0.5260 - val_accuracy: 0.7133\n",
      "Epoch 9/10\n",
      "250/250 [==============================] - 187s 748ms/step - loss: 0.4846 - accuracy: 0.7649 - val_loss: 0.7627 - val_accuracy: 0.7458\n",
      "Epoch 10/10\n",
      "250/250 [==============================] - 156s 625ms/step - loss: 0.4726 - accuracy: 0.7809 - val_loss: 0.3901 - val_accuracy: 0.7238\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x7fae8a8bdda0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit_generator(x_train,\n",
    "                         steps_per_epoch = 250,\n",
    "                         epochs = 10,\n",
    "                         validation_data = x_test,\n",
    "                         validation_steps = 63)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(\"mymodel.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import load_model\n",
    "import numpy as np\n",
    "import cv2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# model = load_model('mymodel.h5')\n",
    "# model.compile(loss='categorical_crossentropy',\n",
    "#                       optimizer='adam',\n",
    "#                       metrics=['accuracy'])\n",
    "\n",
    "from skimage.transform import resize\n",
    "\n",
    "def detect(frame):\n",
    "    try:\n",
    "        img = resize(frame,(64,64))\n",
    "        img = np.expand_dims(img,axis=0)\n",
    "        if(np.max(img)>1):\n",
    "            img = img/255.0\n",
    "        prediction = model.predict(img)\n",
    "        print(prediction)\n",
    "        prediction = model.predict_classes(img)\n",
    "        print(prediction)\n",
    "    except AttributeError:\n",
    "        print(\"shape not found\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.9486049]]\n",
      "[[1]]\n"
     ]
    }
   ],
   "source": [
    "frame=cv2.imread(\"dog.jpg\")\n",
    "data = detect(frame)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}