USC_CNN_v3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Urban Sound Classifier using CNN v2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'2.2.0'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# draw\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('ggplot')\n",
    "# basic handling\n",
    "import os\n",
    "import glob\n",
    "import pickle\n",
    "import numpy as np\n",
    "# audio\n",
    "import librosa\n",
    "import librosa.display\n",
    "import IPython.display\n",
    "# normalization\n",
    "import sklearn\n",
    "# nn\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, MaxPooling2D, AveragePooling2D, BatchNormalization\n",
    "from keras.layers.convolutional import Conv2D,MaxPooling2D\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.utils import to_categorical\n",
    "from keras.callbacks import TensorBoard\n",
    "keras.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_history(history):\n",
    "    print(history.history.keys())\n",
    "    fig = plt.figure(figsize=(20,5))\n",
    "    plt.subplot(121)\n",
    "    plt.plot(history.history['acc'])\n",
    "    plt.plot(history.history['val_acc'])\n",
    "    plt.title('model accuracy')\n",
    "    plt.ylabel('accuracy')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.legend(['train', 'test'], loc='upper left')\n",
    "    plt.subplot(122)\n",
    "    plt.plot(history.history['loss'])\n",
    "    plt.plot(history.history['val_loss'])\n",
    "    plt.title('model loss')\n",
    "    plt.ylabel('loss')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.legend(['train', 'test'], loc='lower left')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = pickle.load(open('./train_x.dat', 'rb'))\n",
    "train_y = pickle.load(open('./train_y.dat', 'rb'))\n",
    "val_x = pickle.load(open('./val_x.dat', 'rb'))\n",
    "val_y = pickle.load(open('./val_y.dat', 'rb'))\n",
    "test_x = pickle.load(open('./test_x.dat', 'rb'))\n",
    "test_y = pickle.load(open('./test_y.dat', 'rb'))\n",
    "\n",
    "train_x = train_x.reshape(train_x.shape[0],train_x.shape[1],train_x.shape[2],1)\n",
    "val_x = val_x.reshape(val_x.shape[0],val_x.shape[1],val_x.shape[2],1)\n",
    "test_x = test_x.reshape(test_x.shape[0],test_x.shape[1],test_x.shape[2],1)\n",
    "\n",
    "train_y = to_categorical(train_y)\n",
    "val_y = to_categorical(val_y)\n",
    "test_y = to_categorical(test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "________________________________________________________________________________\n",
      "Layer (type)                        Output Shape                    Param #     \n",
      "================================================================================\n",
      "conv2d_1 (Conv2D)                   (None, 40, 87, 32)              320         \n",
      "________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)      (None, 20, 43, 32)              0           \n",
      "________________________________________________________________________________\n",
      "dropout_1 (Dropout)                 (None, 20, 43, 32)              0           \n",
      "________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)                   (None, 20, 22, 64)              18496       \n",
      "________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)                   (None, 20, 22, 64)              36928       \n",
      "________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNormali (None, 20, 22, 64)              256         \n",
      "________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)      (None, 10, 11, 64)              0           \n",
      "________________________________________________________________________________\n",
      "dropout_2 (Dropout)                 (None, 10, 11, 64)              0           \n",
      "________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)                   (None, 10, 11, 128)             73856       \n",
      "________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)                   (None, 10, 11, 128)             147584      \n",
      "________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNormali (None, 10, 11, 128)             512         \n",
      "________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)      (None, 5, 5, 128)               0           \n",
      "________________________________________________________________________________\n",
      "dropout_3 (Dropout)                 (None, 5, 5, 128)               0           \n",
      "________________________________________________________________________________\n",
      "flatten_1 (Flatten)                 (None, 3200)                    0           \n",
      "________________________________________________________________________________\n",
      "dense_1 (Dense)                     (None, 512)                     1638912     \n",
      "________________________________________________________________________________\n",
      "dropout_4 (Dropout)                 (None, 512)                     0           \n",
      "________________________________________________________________________________\n",
      "dense_2 (Dense)                     (None, 10)                      5130        \n",
      "================================================================================\n",
      "Total params: 1,921,994\n",
      "Trainable params: 1,921,610\n",
      "Non-trainable params: 384\n",
      "________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Conv2D(32,(3,3),strides=(1,2),padding='same',activation='relu',kernel_initializer='uniform',input_shape = train_x.shape[1:]))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Conv2D(64,(3,3),strides=(1,2),padding='same',activation='relu',kernel_initializer='uniform'))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(Conv2D(64,(3,3),strides=(1,1),padding='same',activation='relu',kernel_initializer='uniform'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Conv2D(128,(3,3),strides=(1,1),padding='same',activation='relu',kernel_initializer='uniform'))\n",
    "# model.add(BatchNormalization())\n",
    "model.add(Conv2D(128,(3,3),strides=(1,1),padding='same',activation='relu',kernel_initializer='uniform'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512,activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(10,activation='softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy',optimizer='Adam',metrics=['accuracy'])\n",
    "model.summary(line_length=80)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2148 samples, validate on 538 samples\n",
      "Epoch 1/20\n",
      "2148/2148 [==============================] - 9s 4ms/step - loss: 3.5615 - acc: 0.1904 - val_loss: 9.3594 - val_acc: 0.0428\n",
      "Epoch 2/20\n",
      "2148/2148 [==============================] - 5s 2ms/step - loss: 2.1878 - acc: 0.3613 - val_loss: 2.5361 - val_acc: 0.1784\n",
      "Epoch 3/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 1.8209 - acc: 0.4288 - val_loss: 4.0209 - val_acc: 0.2026\n",
      "Epoch 4/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 1.5542 - acc: 0.4758 - val_loss: 2.8101 - val_acc: 0.1561\n",
      "Epoch 5/20\n",
      "2148/2148 [==============================] - 5s 2ms/step - loss: 1.2380 - acc: 0.5345 - val_loss: 1.9501 - val_acc: 0.3717\n",
      "Epoch 6/20\n",
      "2148/2148 [==============================] - 5s 3ms/step - loss: 0.9872 - acc: 0.6192 - val_loss: 5.0040 - val_acc: 0.2732\n",
      "Epoch 7/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.8867 - acc: 0.6583 - val_loss: 2.8394 - val_acc: 0.4257\n",
      "Epoch 8/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.7319 - acc: 0.7309 - val_loss: 2.4421 - val_acc: 0.4201\n",
      "Epoch 9/20\n",
      "2148/2148 [==============================] - 5s 3ms/step - loss: 0.6819 - acc: 0.7579 - val_loss: 2.0701 - val_acc: 0.5019\n",
      "Epoch 10/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.6114 - acc: 0.7900 - val_loss: 3.6235 - val_acc: 0.3067\n",
      "Epoch 11/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.5266 - acc: 0.8115 - val_loss: 2.0095 - val_acc: 0.4944\n",
      "Epoch 12/20\n",
      "2148/2148 [==============================] - 5s 2ms/step - loss: 0.4894 - acc: 0.8352 - val_loss: 4.2644 - val_acc: 0.3253\n",
      "Epoch 13/20\n",
      "2148/2148 [==============================] - 5s 3ms/step - loss: 0.4423 - acc: 0.8454 - val_loss: 2.4608 - val_acc: 0.3959\n",
      "Epoch 14/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.3679 - acc: 0.8659 - val_loss: 4.9094 - val_acc: 0.2416\n",
      "Epoch 15/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.3709 - acc: 0.8706 - val_loss: 2.4301 - val_acc: 0.4349\n",
      "Epoch 16/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.3064 - acc: 0.8901 - val_loss: 3.0771 - val_acc: 0.4758\n",
      "Epoch 17/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.2446 - acc: 0.9143 - val_loss: 3.4169 - val_acc: 0.5019\n",
      "Epoch 18/20\n",
      "2148/2148 [==============================] - 5s 2ms/step - loss: 0.2555 - acc: 0.9055 - val_loss: 4.0350 - val_acc: 0.4089\n",
      "Epoch 19/20\n",
      "2148/2148 [==============================] - 7s 3ms/step - loss: 0.1903 - acc: 0.9278 - val_loss: 3.3114 - val_acc: 0.4442\n",
      "Epoch 20/20\n",
      "2148/2148 [==============================] - 6s 3ms/step - loss: 0.1584 - acc: 0.9460 - val_loss: 3.0003 - val_acc: 0.4888\n",
      "dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# callbacks=[TensorBoard(log_dir='./log')]\n",
    "history = model.fit(train_x, train_y, \n",
    "                    epochs=20, batch_size=128, \n",
    "                    validation_split=0.2)\n",
    "show_history(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "837/837 [==============================] - 0s 389us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[2.5665002526490635, 0.5364396655431358]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(test_x,test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}