VGG16_Transferlearning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import required libraries - we do need the models, flatten, dense, input layers \n",
    "import numpy as np \n",
    "import pandas as pd\n",
    "from glob import glob \n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.keras.layers import Input, Lambda, Dense, Flatten\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.applications.vgg16 import VGG16\n",
    "from tensorflow.keras.applications.vgg19 import VGG19\n",
    "from tensorflow.keras.preprocessing import image\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.models import Sequential"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 3458 images belonging to 5 classes.\n",
      "Found 430 images belonging to 5 classes.\n",
      "Found 435 images belonging to 5 classes.\n"
     ]
    }
   ],
   "source": [
    "# Path for train, validation and test datasets\n",
    "train_path = 'Datasets/train'\n",
    "valid_path = 'Datasets/test'\n",
    "test_path = 'Datasets/val' \n",
    "\n",
    "IMAGE_SIZE = [224, 224] #Default image size for VGG16 \n",
    "\n",
    "folders = glob('Datasets/train/*') #Get number of classes\n",
    "\n",
    "\n",
    "# ImageDataGenerator can help perform augumentation on existing images. This way, we get more diverse train set.\n",
    "train_datagen = ImageDataGenerator(rescale = 1./255, shear_range = 0.2, zoom_range = 0.2, horizontal_flip = True)\n",
    "validation_datagen = ImageDataGenerator(rescale = 1./255)\n",
    "test_datagen = ImageDataGenerator(rescale = 1./255)\n",
    "\n",
    "#Through flow_from_directory - we create an array of images that can be used for training. \n",
    "training_set = train_datagen.flow_from_directory('Datasets/train',\n",
    "                                                 target_size = (224, 224),\n",
    "                                                 batch_size = 64,\n",
    "                                                 class_mode = 'categorical')\n",
    "\n",
    "validation_set = validation_datagen.flow_from_directory('Datasets/val',\n",
    "                                                 target_size = (224, 224),\n",
    "                                                 batch_size = 64,\n",
    "                                                 class_mode = 'categorical')\n",
    "\n",
    "test_set = test_datagen.flow_from_directory('Datasets/test',\n",
    "                                            target_size = (224, 224),\n",
    "                                            batch_size = 32,\n",
    "                                            class_mode = 'categorical')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"functional_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 224, 224, 3)]     0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 25088)             0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 5)                 125445    \n",
      "=================================================================\n",
      "Total params: 14,840,133\n",
      "Trainable params: 125,445\n",
      "Non-trainable params: 14,714,688\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Create a VGG16 model, and removing the last layer that is classifying 1000 images. This will be replaced with images classes we have. \n",
    "vgg = VGG16(input_shape=IMAGE_SIZE + [3], weights='imagenet', include_top=False) #Training with Imagenet weights\n",
    "\n",
    "# Use this line for VGG19 network. Create a VGG19 model, and removing the last layer that is classifying 1000 images. This will be replaced with images classes we have. \n",
    "#vgg = VGG19(input_shape=IMAGE_SIZE + [3], weights='imagenet', include_top=False) \n",
    "\n",
    "# This sets the base that the layers are not trainable. If we'd want to train the layers with custom data, these two lines can be ommitted. \n",
    "for layer in vgg.layers:\n",
    "  layer.trainable = False\n",
    "\n",
    "x = Flatten()(vgg.output) #Output obtained on vgg16 is now flattened. \n",
    "prediction = Dense(len(folders), activation='softmax')(x) # We have 5 classes, and so, the prediction is being done on len(folders) - 5 classes\n",
    "\n",
    "#Creating model object \n",
    "model = Model(inputs=vgg.input, outputs=prediction)\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "55/55 [==============================] - 49s 888ms/step - loss: 0.9809 - accuracy: 0.6382 - val_loss: 0.5901 - val_accuracy: 0.7767\n",
      "Epoch 2/20\n",
      "55/55 [==============================] - 40s 731ms/step - loss: 0.5321 - accuracy: 0.8039 - val_loss: 0.5046 - val_accuracy: 0.8209\n",
      "Epoch 3/20\n",
      "55/55 [==============================] - 40s 731ms/step - loss: 0.4723 - accuracy: 0.8317 - val_loss: 0.4933 - val_accuracy: 0.8209\n",
      "Epoch 4/20\n",
      "55/55 [==============================] - 41s 746ms/step - loss: 0.3879 - accuracy: 0.8623 - val_loss: 0.4989 - val_accuracy: 0.8209\n",
      "Epoch 5/20\n",
      "55/55 [==============================] - 43s 780ms/step - loss: 0.4052 - accuracy: 0.8623 - val_loss: 0.4580 - val_accuracy: 0.8302\n",
      "Epoch 6/20\n",
      "55/55 [==============================] - 46s 835ms/step - loss: 0.3074 - accuracy: 0.8895 - val_loss: 0.4751 - val_accuracy: 0.8395\n",
      "Epoch 7/20\n",
      "55/55 [==============================] - 42s 772ms/step - loss: 0.2324 - accuracy: 0.9263 - val_loss: 0.4777 - val_accuracy: 0.8442\n",
      "Epoch 8/20\n",
      "55/55 [==============================] - 41s 753ms/step - loss: 0.2215 - accuracy: 0.9315 - val_loss: 0.4626 - val_accuracy: 0.8233\n",
      "Epoch 9/20\n",
      "55/55 [==============================] - 39s 716ms/step - loss: 0.1954 - accuracy: 0.9422 - val_loss: 0.4676 - val_accuracy: 0.8302\n",
      "Epoch 10/20\n",
      "55/55 [==============================] - 37s 675ms/step - loss: 0.1711 - accuracy: 0.9523 - val_loss: 0.5003 - val_accuracy: 0.8326\n",
      "Epoch 11/20\n",
      "55/55 [==============================] - 37s 670ms/step - loss: 0.1664 - accuracy: 0.9503 - val_loss: 0.4782 - val_accuracy: 0.8186\n",
      "Epoch 12/20\n",
      "55/55 [==============================] - 36s 653ms/step - loss: 0.1379 - accuracy: 0.9647 - val_loss: 0.5253 - val_accuracy: 0.8209\n",
      "Epoch 13/20\n",
      "55/55 [==============================] - 36s 657ms/step - loss: 0.1372 - accuracy: 0.9639 - val_loss: 0.4777 - val_accuracy: 0.8302\n",
      "Epoch 14/20\n",
      "55/55 [==============================] - 37s 679ms/step - loss: 0.1247 - accuracy: 0.9691 - val_loss: 0.5150 - val_accuracy: 0.8209\n",
      "Epoch 15/20\n",
      "55/55 [==============================] - 37s 678ms/step - loss: 0.1408 - accuracy: 0.9575 - val_loss: 0.5651 - val_accuracy: 0.8186\n",
      "Epoch 16/20\n",
      "55/55 [==============================] - 36s 651ms/step - loss: 0.1177 - accuracy: 0.9636 - val_loss: 0.5064 - val_accuracy: 0.8209\n",
      "Epoch 17/20\n",
      "55/55 [==============================] - 36s 654ms/step - loss: 0.0973 - accuracy: 0.9746 - val_loss: 0.5191 - val_accuracy: 0.8372\n",
      "Epoch 18/20\n",
      "55/55 [==============================] - 36s 652ms/step - loss: 0.1082 - accuracy: 0.9691 - val_loss: 0.5489 - val_accuracy: 0.8116\n",
      "Epoch 19/20\n",
      "55/55 [==============================] - 36s 652ms/step - loss: 0.1003 - accuracy: 0.9737 - val_loss: 0.5018 - val_accuracy: 0.8395\n",
      "Epoch 20/20\n",
      "55/55 [==============================] - 36s 653ms/step - loss: 0.0904 - accuracy: 0.9769 - val_loss: 0.5375 - val_accuracy: 0.8186\n"
     ]
    }
   ],
   "source": [
    "#Compile the model \n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) \n",
    "history = model.fit(training_set, validation_data=validation_set, epochs=20, batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# loss\n",
    "plt.plot(history.history['loss'], label='train loss')\n",
    "plt.plot(history.history['val_loss'], label='val loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.savefig('LossVal_loss')\n",
    "\n",
    "# accuracies\n",
    "plt.plot(history.history['accuracy'], label='train acc')\n",
    "plt.plot(history.history['val_accuracy'], label='val acc')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.savefig('AccVal_acc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import load_model\n",
    "model.save('FlowerClassification_VGG16.h5')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "KerasEnv",
   "language": "python",
   "name": "kerasenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}