preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "7c524989",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "data_gen = ImageDataGenerator(rotation_range=30, rescale=1./255, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "8293d819",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 14140 images belonging to 101 classes.\n",
      "Found 6060 images belonging to 101 classes.\n"
     ]
    }
   ],
   "source": [
    "image_size = (235, 235)\n",
    "train_gen = data_gen.flow_from_directory('train', target_size=image_size, shuffle=True, seed=42) # class_mode default categorical\n",
    "test_gen = data_gen.flow_from_directory('test', target_size=image_size, shuffle=True, seed=42) # default batch_size is 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "d73c7a29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  0,   0,   0, ..., 100, 100, 100], dtype=int32)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_gen.labels\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "bb4b68ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.applications import InceptionV3\n",
    "\n",
    "input_shape = (235, 235, 3) # 235, 235 stands for w and h. The 3 stands for the three channels for RGB\n",
    "inception_model = InceptionV3(weights=\"imagenet\", input_shape=input_shape, include_top=False)\n",
    "\n",
    "inception_model.trainable = False # freeze weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "461d556d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " inception_v3 (Functional)   (None, 6, 6, 2048)        21802784  \n",
      "                                                                 \n",
      " global_average_pooling2d_2   (None, 2048)             0         \n",
      " (GlobalAveragePooling2D)                                        \n",
      "                                                                 \n",
      " dense_10 (Dense)            (None, 128)               262272    \n",
      "                                                                 \n",
      " dense_11 (Dense)            (None, 128)               16512     \n",
      "                                                                 \n",
      " dense_12 (Dense)            (None, 128)               16512     \n",
      "                                                                 \n",
      " dense_13 (Dense)            (None, 128)               16512     \n",
      "                                                                 \n",
      " dropout_2 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_14 (Dense)            (None, 101)               13029     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 22,127,621\n",
      "Trainable params: 324,837\n",
      "Non-trainable params: 21,802,784\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Activation, Dropout, Conv2D, MaxPool2D, GlobalAveragePooling2D\n",
    "\n",
    "model = Sequential()\n",
    "model.add(inception_model)\n",
    "model.add(GlobalAveragePooling2D()) \n",
    "model.add(Dense(128, activation=\"relu\")) # 64 units\n",
    "model.add(Dense(128, activation=\"relu\")) # added another layer\n",
    "model.add(Dense(128, activation=\"relu\")) # added a third layer\n",
    "model.add(Dense(128, activation=\"relu\")) # added a fourth layer\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(101, activation='softmax'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "f15b746f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.optimizers import Adam\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "f9d68149",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "\n",
    "earlystopping = EarlyStopping(min_delta=0.001, patience=5, restore_best_weights=True)\n",
    "\n",
    "checkpoint = ModelCheckpoint(filepath='food_model.hdf5', verbose=1, save_best_only=True, save_weights_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "335caff1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-04-11 20:47:32.054952: 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": [
      "442/442 [==============================] - ETA: 0s - loss: 3.7780 - accuracy: 0.1286"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-04-11 20:49:17.393769: 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": [
      "\n",
      "Epoch 1: val_loss improved from inf to 3.06774, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 153s 341ms/step - loss: 3.7780 - accuracy: 0.1286 - val_loss: 3.0677 - val_accuracy: 0.2353\n",
      "Epoch 2/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 3.0029 - accuracy: 0.2535\n",
      "Epoch 2: val_loss improved from 3.06774 to 2.77040, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 150s 339ms/step - loss: 3.0029 - accuracy: 0.2535 - val_loss: 2.7704 - val_accuracy: 0.2988\n",
      "Epoch 3/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.7309 - accuracy: 0.3125\n",
      "Epoch 3: val_loss improved from 2.77040 to 2.64810, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 152s 344ms/step - loss: 2.7309 - accuracy: 0.3125 - val_loss: 2.6481 - val_accuracy: 0.3314\n",
      "Epoch 4/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.5596 - accuracy: 0.3471\n",
      "Epoch 4: val_loss improved from 2.64810 to 2.54843, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 150s 340ms/step - loss: 2.5596 - accuracy: 0.3471 - val_loss: 2.5484 - val_accuracy: 0.3520\n",
      "Epoch 5/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.4133 - accuracy: 0.3837\n",
      "Epoch 5: val_loss improved from 2.54843 to 2.52742, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 149s 338ms/step - loss: 2.4133 - accuracy: 0.3837 - val_loss: 2.5274 - val_accuracy: 0.3644\n",
      "Epoch 6/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.3240 - accuracy: 0.4011\n",
      "Epoch 6: val_loss improved from 2.52742 to 2.38933, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 148s 336ms/step - loss: 2.3240 - accuracy: 0.4011 - val_loss: 2.3893 - val_accuracy: 0.3960\n",
      "Epoch 7/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.2472 - accuracy: 0.4161\n",
      "Epoch 7: val_loss did not improve from 2.38933\n",
      "442/442 [==============================] - 148s 335ms/step - loss: 2.2472 - accuracy: 0.4161 - val_loss: 2.3900 - val_accuracy: 0.4013\n",
      "Epoch 8/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.1872 - accuracy: 0.4322\n",
      "Epoch 8: val_loss improved from 2.38933 to 2.38847, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 148s 336ms/step - loss: 2.1872 - accuracy: 0.4322 - val_loss: 2.3885 - val_accuracy: 0.4045\n",
      "Epoch 9/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.1270 - accuracy: 0.4446\n",
      "Epoch 9: val_loss improved from 2.38847 to 2.33956, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 149s 337ms/step - loss: 2.1270 - accuracy: 0.4446 - val_loss: 2.3396 - val_accuracy: 0.4172\n",
      "Epoch 10/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.0728 - accuracy: 0.4558\n",
      "Epoch 10: val_loss did not improve from 2.33956\n",
      "442/442 [==============================] - 148s 334ms/step - loss: 2.0728 - accuracy: 0.4558 - val_loss: 2.3975 - val_accuracy: 0.4063\n",
      "Epoch 11/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 2.0094 - accuracy: 0.4728\n",
      "Epoch 11: val_loss improved from 2.33956 to 2.32874, saving model to food_model.hdf5\n",
      "442/442 [==============================] - 149s 337ms/step - loss: 2.0094 - accuracy: 0.4728 - val_loss: 2.3287 - val_accuracy: 0.4185\n",
      "Epoch 12/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 1.9657 - accuracy: 0.4821\n",
      "Epoch 12: val_loss did not improve from 2.32874\n",
      "442/442 [==============================] - 149s 337ms/step - loss: 1.9657 - accuracy: 0.4821 - val_loss: 2.3695 - val_accuracy: 0.4008\n",
      "Epoch 13/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 1.9277 - accuracy: 0.4877\n",
      "Epoch 13: val_loss did not improve from 2.32874\n",
      "442/442 [==============================] - 149s 337ms/step - loss: 1.9277 - accuracy: 0.4877 - val_loss: 2.4030 - val_accuracy: 0.4152\n",
      "Epoch 14/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 1.9020 - accuracy: 0.4954\n",
      "Epoch 14: val_loss did not improve from 2.32874\n",
      "442/442 [==============================] - 149s 338ms/step - loss: 1.9020 - accuracy: 0.4954 - val_loss: 2.3620 - val_accuracy: 0.4182\n",
      "Epoch 15/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 1.8568 - accuracy: 0.5013\n",
      "Epoch 15: val_loss did not improve from 2.32874\n",
      "442/442 [==============================] - 149s 336ms/step - loss: 1.8568 - accuracy: 0.5013 - val_loss: 2.3809 - val_accuracy: 0.4149\n",
      "Epoch 16/30\n",
      "442/442 [==============================] - ETA: 0s - loss: 1.8378 - accuracy: 0.5072\n",
      "Epoch 16: val_loss did not improve from 2.32874\n",
      "442/442 [==============================] - 151s 341ms/step - loss: 1.8378 - accuracy: 0.5072 - val_loss: 2.3860 - val_accuracy: 0.4186\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_gen, validation_data=test_gen, epochs=30, verbose=1, callbacks=[checkpoint, earlystopping])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "c61a3e81",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights('food_model.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "674bdfce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "190/190 [==============================] - 45s 236ms/step\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "output = np.argmax(model.predict(test_gen), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "7123c8f6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[44  6 90 ... 50 72 13]\n",
      "0.011386138613861386\n"
     ]
    }
   ],
   "source": [
    "print(output)\n",
    "print(np.mean(output == test_gen.labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "b8ca5be4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from tensorflow import keras\n",
    "\n",
    "history_df = pd.DataFrame(history.history)\n",
    "#history_df = keras.models.load_model(\"food_model.hdf5\")\n",
    "history_df.loc[:,['loss', 'val_loss']].plot()\n",
    "history_df.loc[:,['accuracy', 'val_accuracy']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "d8d33672",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4186468720436096\n"
     ]
    }
   ],
   "source": [
    "print(max(history_df['val_accuracy']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26420b87",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}