CAT vs DOGS - SImple Classifier.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D\n",
    "from keras.layers import MaxPooling2D\n",
    "from keras.layers import Flatten\n",
    "from keras.layers import Dense\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create the Convolutional Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "model=Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Conv2D(filters=32,kernel_size=3,activation='relu',input_shape=(64,64,3)))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(MaxPooling2D(pool_size=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Conv2D(filters=32,kernel_size=3,activation='relu',))\n",
    "model.add(MaxPooling2D(pool_size=2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Conv2D(filters=32,kernel_size=3,activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Conv2D(filters=32,kernel_size=3,activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Add a flatten layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Flatten())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary of the CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 62, 62, 32)        896       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 31, 31, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 29, 29, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 12, 12, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 6, 6, 32)          0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 4, 4, 32)          9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 2, 2, 32)          0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 128)               0         \n",
      "=================================================================\n",
      "Total params: 28,640\n",
      "Trainable params: 28,640\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create an Artificial Neural Network on top of the CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))\n",
    "model.add(Dense(units=512,activation='relu',kernel_initializer='uniform'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Dense(units=1,activation='sigmoid',kernel_initializer='uniform'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',loss=\"binary_crossentropy\",metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary of the overall Network (CNN+ANN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 62, 62, 32)        896       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 31, 31, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 29, 29, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 12, 12, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 6, 6, 32)          0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 4, 4, 32)          9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 2, 2, 32)          0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 512)               66048     \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 513       \n",
      "=================================================================\n",
      "Total params: 1,933,793\n",
      "Trainable params: 1,933,793\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Image Augmentation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_datagen = ImageDataGenerator(\n",
    "        rescale=1./255,\n",
    "        shear_range=0.2,\n",
    "        zoom_range=0.2,\n",
    "        horizontal_flip=True,)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_datagen = ImageDataGenerator(rescale=1./255)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set the Directory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### \"YOUR FILE PATH\" is the local path of your machine where you have set up your folders for training and test sets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2000 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "train_set = train_datagen.flow_from_directory (\n",
    "    './training_set',\\\n",
    "    target_size=(64, 64),\\\n",
    "    batch_size=10,\\\n",
    "    class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1000 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "test_set = test_datagen.flow_from_directory(\n",
    "    './test_set',\\\n",
    "        target_size=(64, 64),\\\n",
    "        batch_size=10,\\\n",
    "        class_mode='binary')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fit the Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "### steps_per_epoch = Total Number of images in the training set\n",
    "### validation_steps = Total number of Images in the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "200"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "20"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "steps_per_epoch = len(train_set)//10\n",
    "steps_per_epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "20/20 [==============================] - 1s 66ms/step - loss: 0.6931 - accuracy: 0.5050 - val_loss: 0.6928 - val_accuracy: 0.5500\n",
      "Epoch 2/10\n",
      "20/20 [==============================] - 1s 65ms/step - loss: 0.6933 - accuracy: 0.4950 - val_loss: 0.6926 - val_accuracy: 0.5600\n",
      "Epoch 3/10\n",
      "20/20 [==============================] - 1s 64ms/step - loss: 0.6931 - accuracy: 0.5100 - val_loss: 0.6936 - val_accuracy: 0.4400\n",
      "Epoch 4/10\n",
      "20/20 [==============================] - 1s 63ms/step - loss: 0.6932 - accuracy: 0.5100 - val_loss: 0.6934 - val_accuracy: 0.4800\n",
      "Epoch 5/10\n",
      "20/20 [==============================] - 1s 63ms/step - loss: 0.6934 - accuracy: 0.4750 - val_loss: 0.6927 - val_accuracy: 0.6200\n",
      "Epoch 6/10\n",
      "20/20 [==============================] - 1s 61ms/step - loss: 0.6932 - accuracy: 0.4850 - val_loss: 0.6932 - val_accuracy: 0.4400\n",
      "Epoch 7/10\n",
      "20/20 [==============================] - 1s 64ms/step - loss: 0.6931 - accuracy: 0.5350 - val_loss: 0.6931 - val_accuracy: 0.5300\n",
      "Epoch 8/10\n",
      "20/20 [==============================] - 1s 62ms/step - loss: 0.6933 - accuracy: 0.4600 - val_loss: 0.6931 - val_accuracy: 0.5200\n",
      "Epoch 9/10\n",
      "20/20 [==============================] - 1s 62ms/step - loss: 0.6932 - accuracy: 0.5050 - val_loss: 0.6934 - val_accuracy: 0.4400\n",
      "Epoch 10/10\n",
      "20/20 [==============================] - 1s 62ms/step - loss: 0.6930 - accuracy: 0.5250 - val_loss: 0.6928 - val_accuracy: 0.5500\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1d4d6185400>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit_generator(\n",
    "        train_set,\n",
    "        steps_per_epoch=20,\n",
    "        epochs=10,\n",
    "        validation_data=test_set,\n",
    "        validation_steps=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting a New Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing import image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Target size is 64x64 as out CNN inputs the image size as 64x64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_image=image.load_img('./test_set/dogs/dog.2000.jpg',target_size=(64,64))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=64x64 at 0x1D4DA97FEE0>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Change the image to  numpy array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_image=image.img_to_array(new_image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_image.ndim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(new_image)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The input needs to be in 4 dimesnion. 4th Dim represents the batch size so add one more dimesion using the expand_dims function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_image=np.expand_dims(new_image,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_image.ndim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 64, 64, 3)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_image.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Predict the image and store it in a variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "result=model.predict(new_image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.50244063]], dtype=float32)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cats': 0, 'dogs': 1}"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_set.class_indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.ndim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.50244063"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result[0][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This is a Dog\n"
     ]
    }
   ],
   "source": [
    "result=model.predict(new_image)\n",
    "if result[0][0]>0.5:\n",
    "    predict=print('This is a Dog')\n",
    "else:\n",
    "    predict=print('This is a Cat')\n",
    "    \n",
    "    \n"
   ]
  },
  {
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}