{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from keras.preprocessing import image\n",
    "import matplotlib.pyplot as plt \n",
    "import numpy as np\n",
    "from keras.utils.np_utils import to_categorical\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dropout,Conv2D,Flatten,Dense,MaxPooling2D,BatchNormalization\n",
    "from keras.models import load_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 39562 images belonging to 2 classes.\n",
      "Found 9890 images belonging to 2 classes.\n",
      "steps_per_epoch 1236, validation_steps 309\n"
     ]
    }
   ],
   "source": [
    "batch_size=32\n",
    "\n",
    "train_datagen = image.ImageDataGenerator(\n",
    "  rescale=1./255,\n",
    "  rotation_range=30,\n",
    "  validation_split=0.2,\n",
    "  width_shift_range=0.1,\n",
    "  height_shift_range=0.1,\n",
    "  shear_range=0.1,\n",
    "  fill_mode='wrap'\n",
    ")\n",
    "def generate(subset='training', data_dir='../data/eye', shuffle=False, target_size=(24,24), color_mode='grayscale', class_mode='categorical'):\n",
    "  return train_datagen.flow_from_directory(data_dir, batch_size=batch_size, target_size=target_size, color_mode=color_mode, class_mode=class_mode, shuffle=shuffle, subset=subset)\n",
    "train_data = generate(subset='training', shuffle=True)\n",
    "validation_data = generate(subset='validation')\n",
    "\n",
    "steps_per_epoch = len(train_data.classes) // batch_size\n",
    "validation_steps= len(validation_data.classes) // batch_size\n",
    "print(f\"steps_per_epoch {steps_per_epoch}, validation_steps {validation_steps}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def CNN():\n",
    "    model = Sequential([\n",
    "        Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(24,24,1)),\n",
    "        MaxPooling2D(pool_size=(2,2),strides=1),\n",
    "        Conv2D(32,(3,3),activation='relu'),\n",
    "        MaxPooling2D(pool_size=(2,2),strides=1),\n",
    " \n",
    "        Conv2D(64, (3, 3), activation='relu'),\n",
    "        MaxPooling2D(pool_size=(2,2), strides=1),\n",
    "\n",
    "        Dropout(0.25),\n",
    "   \n",
    "        Flatten(),\n",
    "    \n",
    "        Dense(128, activation='relu'),\n",
    "    \n",
    "        Dropout(0.5),\n",
    "    \n",
    "        Dense(2, activation='softmax')\n",
    "    ])\n",
    "    \n",
    "    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy']) \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metal device set to: Apple M1 Pro\n",
      "\n",
      "systemMemory: 32.00 GB\n",
      "maxCacheSize: 10.67 GB\n",
      "\n",
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 22, 22, 32)        320       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 21, 21, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 19, 19, 32)        9248      \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 18, 18, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 16, 16, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 15, 15, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 15, 15, 64)        0         \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 14400)             0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 128)               1843328   \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 2)                 258       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1,871,650\n",
      "Trainable params: 1,871,650\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-05-14 13:45:06.284204: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:305] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2022-05-14 13:45:06.284338: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:271] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    }
   ],
   "source": [
    "CNN().summary()\n",
    "model = CNN()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-05-14 13:45:09.482774: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n",
      "2022-05-14 13:45:09.770662: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1236/1236 [==============================] - ETA: 0s - loss: 0.2562 - accuracy: 0.8935"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-05-14 13:45:26.939899: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1236/1236 [==============================] - 22s 15ms/step - loss: 0.2562 - accuracy: 0.8935 - val_loss: 0.2637 - val_accuracy: 0.9154\n",
      "Epoch 2/15\n",
      "1236/1236 [==============================] - 18s 14ms/step - loss: 0.1293 - accuracy: 0.9520 - val_loss: 0.2223 - val_accuracy: 0.9199\n",
      "Epoch 3/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0960 - accuracy: 0.9648 - val_loss: 0.1709 - val_accuracy: 0.9364\n",
      "Epoch 4/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0761 - accuracy: 0.9719 - val_loss: 0.1932 - val_accuracy: 0.9344\n",
      "Epoch 5/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0672 - accuracy: 0.9765 - val_loss: 0.1815 - val_accuracy: 0.9328\n",
      "Epoch 6/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0585 - accuracy: 0.9790 - val_loss: 0.1783 - val_accuracy: 0.9384\n",
      "Epoch 7/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0532 - accuracy: 0.9814 - val_loss: 0.1681 - val_accuracy: 0.9447\n",
      "Epoch 8/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0505 - accuracy: 0.9823 - val_loss: 0.2100 - val_accuracy: 0.9315\n",
      "Epoch 9/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0468 - accuracy: 0.9838 - val_loss: 0.1515 - val_accuracy: 0.9513\n",
      "Epoch 10/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0453 - accuracy: 0.9844 - val_loss: 0.1636 - val_accuracy: 0.9511\n",
      "Epoch 11/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0441 - accuracy: 0.9845 - val_loss: 0.1591 - val_accuracy: 0.9491\n",
      "Epoch 12/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0405 - accuracy: 0.9855 - val_loss: 0.1706 - val_accuracy: 0.9561\n",
      "Epoch 13/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0405 - accuracy: 0.9863 - val_loss: 0.1721 - val_accuracy: 0.9500\n",
      "Epoch 14/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0401 - accuracy: 0.9866 - val_loss: 0.1812 - val_accuracy: 0.9423\n",
      "Epoch 15/15\n",
      "1236/1236 [==============================] - 17s 14ms/step - loss: 0.0395 - accuracy: 0.9864 - val_loss: 0.2955 - val_accuracy: 0.9441\n"
     ]
    }
   ],
   "source": [
    "history=model.fit(train_data, validation_data=validation_data, epochs=15, steps_per_epoch=steps_per_epoch, validation_steps=validation_steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['accuracy'])\n",
    "plt.title('accuracy')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('val_acuuracy')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model val loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['loss', 'val_loss', 'accuracy', 'val_accuracy'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('../model/extended_weight.h5', overwrite=True)"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "d9dade41cb61ad4dc71015ff9c058e4e51f62a5c87dfb0aca90e14591e928332"
  },
  "kernelspec": {
   "display_name": "Python 3.9.12 ('drowsiness_detector')",
   "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": "3.9.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}