PD_feature_extraction_w_data_augmentation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.7-predicting-house-prices.ipynb\r\n",
      "DLwP 3.3.ipynb\r\n",
      "input\r\n",
      "lost+found\r\n",
      "PD_fast_feature_extraction_wo_data_augmentation.ipynb\r\n",
      "PD_feature_extraction_w_data_augmentation.ipynb\r\n",
      "plant_dataset_02.zip\r\n"
     ]
    }
   ],
   "source": [
    "!ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!rm -rf input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!unzip plant_dataset_02.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test  train  validation\r\n"
     ]
    }
   ],
   "source": [
    "!ls input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras import models\n",
    "from keras import layers\n",
    "\n",
    "from keras.applications.resnet50 import ResNet50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_shape = (300, 300, 3)\n",
    "target_size = (300, 300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base = ResNet50(weights='imagenet', input_shape=input_shape, include_top=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import optimizers\n",
    "\n",
    "model = models.Sequential()\n",
    "model.add(conv_base)\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dense(3, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "resnet50 (Model)             (None, 7, 7, 2048)        23587712  \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 100352)            0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 256)               25690368  \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 3)                 771       \n",
      "=================================================================\n",
      "Total params: 49,278,851\n",
      "Trainable params: 49,225,731\n",
      "Non-trainable params: 53,120\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This is the number of trainable weights  before freezing the conv base: 216\n"
     ]
    }
   ],
   "source": [
    "print('This is the number of trainable weights ',\n",
    "     'before freezing the conv base:', len(model.trainable_weights))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This is the number of trainable weights  after freezing the conv base: 4\n"
     ]
    }
   ],
   "source": [
    "print('This is the number of trainable weights ',\n",
    "     'after freezing the conv base:', len(model.trainable_weights))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1826 images belonging to 3 classes.\n",
      "Found 596 images belonging to 3 classes.\n"
     ]
    }
   ],
   "source": [
    "# make data generators\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1./255,\n",
    "    rotation_range=40,\n",
    "    width_shift_range=0.2,\n",
    "    height_shift_range=0.2,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    fill_mode='nearest')\n",
    "test_datagen = ImageDataGenerator(rescale=1./255)\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    'input/train',\n",
    "    target_size=target_size,\n",
    "    batch_size=20,\n",
    "    class_mode='categorical')\n",
    "validation_generator = test_datagen.flow_from_directory(\n",
    "    'input/validation',\n",
    "    target_size=target_size,\n",
    "    batch_size=20,\n",
    "    class_mode='categorical')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy',\n",
    "             optimizer=optimizers.RMSprop(lr=2e-5),\n",
    "             metrics=['acc'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/PIL/Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n",
      "  'to RGBA images')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 100s - loss: 1.0395 - acc: 0.6679 - val_loss: 4.6732 - val_acc: 0.3343\n",
      "Epoch 2/5\n",
      " - 93s - loss: 0.6269 - acc: 0.7490 - val_loss: 6.2573 - val_acc: 0.3313\n",
      "Epoch 3/5\n",
      " - 94s - loss: 0.5495 - acc: 0.7692 - val_loss: 6.9606 - val_acc: 0.3394\n",
      "Epoch 4/5\n",
      " - 92s - loss: 0.5564 - acc: 0.7833 - val_loss: 7.8633 - val_acc: 0.3313\n",
      "Epoch 5/5\n",
      " - 90s - loss: 0.4779 - acc: 0.8138 - val_loss: 7.8169 - val_acc: 0.3313\n"
     ]
    }
   ],
   "source": [
    "history = model.fit_generator(\n",
    "    train_generator,\n",
    "    steps_per_epoch=100,\n",
    "    epochs=5,\n",
    "    verbose=2,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# s 8:45 e 9:15\n",
    "# s 9:20 e "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "acc = history.history['acc']\n",
    "val_acc = history.history['val_acc']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "epochs = range(1, len(acc) + 1)\n",
    "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "plt.figure()\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_tensorflow_p36",
   "language": "python",
   "name": "conda_tensorflow_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}