VGG19_Transferlearning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "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": 2,
   "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": 3,
   "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_conv4 (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_conv4 (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_conv4 (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: 20,149,829\n",
      "Trainable params: 125,445\n",
      "Non-trainable params: 20,024,384\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": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "55/55 [==============================] - 69s 1s/step - loss: 1.1604 - accuracy: 0.5943 - val_loss: 0.6855 - val_accuracy: 0.7860\n",
      "Epoch 2/20\n",
      "55/55 [==============================] - 77s 1s/step - loss: 0.6387 - accuracy: 0.7739 - val_loss: 0.5416 - val_accuracy: 0.8023\n",
      "Epoch 3/20\n",
      "55/55 [==============================] - 78s 1s/step - loss: 0.5048 - accuracy: 0.8155 - val_loss: 0.6214 - val_accuracy: 0.7837\n",
      "Epoch 4/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.4873 - accuracy: 0.8201 - val_loss: 0.5301 - val_accuracy: 0.8279\n",
      "Epoch 5/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.3779 - accuracy: 0.8687 - val_loss: 0.4548 - val_accuracy: 0.8372\n",
      "Epoch 6/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.3255 - accuracy: 0.8924 - val_loss: 0.5248 - val_accuracy: 0.8000\n",
      "Epoch 7/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.3171 - accuracy: 0.8855 - val_loss: 0.4923 - val_accuracy: 0.8442\n",
      "Epoch 8/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.2996 - accuracy: 0.8904 - val_loss: 0.5202 - val_accuracy: 0.8140\n",
      "Epoch 9/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.2578 - accuracy: 0.9118 - val_loss: 0.4537 - val_accuracy: 0.8233\n",
      "Epoch 10/20\n",
      "55/55 [==============================] - 79s 1s/step - loss: 0.2445 - accuracy: 0.9222 - val_loss: 0.4603 - val_accuracy: 0.8442\n",
      "Epoch 11/20\n",
      "55/55 [==============================] - 81s 1s/step - loss: 0.2821 - accuracy: 0.9031 - val_loss: 0.5500 - val_accuracy: 0.8186\n",
      "Epoch 12/20\n",
      "55/55 [==============================] - 73s 1s/step - loss: 0.2366 - accuracy: 0.9211 - val_loss: 0.5157 - val_accuracy: 0.7977\n",
      "Epoch 13/20\n",
      "55/55 [==============================] - 59s 1s/step - loss: 0.2124 - accuracy: 0.9303 - val_loss: 0.4857 - val_accuracy: 0.8349\n",
      "Epoch 14/20\n",
      "55/55 [==============================] - 59s 1s/step - loss: 0.2572 - accuracy: 0.9115 - val_loss: 0.5049 - val_accuracy: 0.8395\n",
      "Epoch 15/20\n",
      "55/55 [==============================] - 63s 1s/step - loss: 0.1691 - accuracy: 0.9445 - val_loss: 0.5562 - val_accuracy: 0.8326\n",
      "Epoch 16/20\n",
      "55/55 [==============================] - 61s 1s/step - loss: 0.1648 - accuracy: 0.9494 - val_loss: 0.4650 - val_accuracy: 0.8465\n",
      "Epoch 17/20\n",
      "55/55 [==============================] - 57s 1s/step - loss: 0.1412 - accuracy: 0.9621 - val_loss: 0.4757 - val_accuracy: 0.8419\n",
      "Epoch 18/20\n",
      "55/55 [==============================] - 65s 1s/step - loss: 0.1512 - accuracy: 0.9537 - val_loss: 0.5046 - val_accuracy: 0.8326\n",
      "Epoch 19/20\n",
      "55/55 [==============================] - 60s 1s/step - loss: 0.1382 - accuracy: 0.9572 - val_loss: 0.4919 - val_accuracy: 0.8349\n",
      "Epoch 20/20\n",
      "55/55 [==============================] - 60s 1s/step - loss: 0.1125 - accuracy: 0.9702 - val_loss: 0.5060 - val_accuracy: 0.8256\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": 5,
   "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": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import load_model\n",
    "model.save('FlowerClassification_VGG19.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
}