TestMain.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Main file to test the models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "from keras.models import Model, load_model\n",
    "#import audioUtils\n",
    "import numpy as np\n",
    "import librosa\n",
    "import SpeechModels\n",
    "from kapre.time_frequency import Melspectrogram, Spectrogram\n",
    "from kapre.utils import Normalization2D\n",
    "from keras.layers import Add, Dropout, BatchNormalization, Conv2D, Reshape, MaxPooling2D, Dense, LSTM, Bidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load the CNN model\n",
    "modelRootDict = \"TrainedModels\"\n",
    "testRootDict = \"Testing\"\n",
    "_cnn_model_path = os.path.join(\".\",modelRootDict,\"Conv_model.h5\")\n",
    "#_rnn_model_path = os.path.join(\".\",modelRootDict,\"AttRNN_model.h5\")\n",
    "_custom_objects={'Melspectrogram':Melspectrogram(),'Normalization2D':Normalization2D(int_axis=0)}\n",
    "global theCNNModel\n",
    "def __load_model(modelPath):\n",
    "    theModel = load_model(modelPath,_custom_objects)\n",
    "    return theModel\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _read_test_wav(wav_file):\n",
    "    '''\n",
    "    read the wav file and convert wav to numpy array.\n",
    "    '''\n",
    "    testfile = os.path.join(\"..\",'Linzy', testRootDict, wav_file)\n",
    "    print(\"reading the test wav file:\",testfile)\n",
    "    return librosa.load(testfile,16000)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _adjustShape(rawArray):\n",
    "    X = np.empty((1,16000))\n",
    "    if rawArray[0] == 16000:\n",
    "            X = rawArray\n",
    "            #print('Same dim')\n",
    "    elif rawArray.shape[0] > 16000: #bigger\n",
    "            #we can choose any position in curX-self.dim\n",
    "            randPos = np.random.randint(rawArray.shape[0]-16000) \n",
    "            X = rawArray[randPos:randPos+16000]\n",
    "            #print('File dim bigger')\n",
    "    else: #smaller\n",
    "            randPos = np.random.randint(16000-rawArray.shape[0])\n",
    "            X[i,randPos:randPos+rawArray.shape[0]] = rawArray\n",
    "            #print('File dim smaller')\n",
    "    return X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def main():\n",
    "    print(\"Loading the Model...........\")\n",
    "    __theCNNModel = __load_model(_cnn_model_path)\n",
    "    print(\"Reading the test wav file.........\")\n",
    "    y, sr = _read_test_wav(\"Ho1226_8894_0.wav\")\n",
    "    y = _adjustShape(y)#np.transpose(_adjustShape(y))\n",
    "    __theCNNModel.summary()\n",
    "    input_y = y.reshape(1,16000)\n",
    "    print(y.shape)\n",
    "    #print(\"y's shape is {0} and {1}\".format(y.shape, sr))\n",
    "    y_result = __theCNNModel.predict(input_y)\n",
    "    #print(\"The predicting result is:\",y_result)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading the Model...........\n",
      "Reading the test wav file.........\n",
      "reading the test wav file: ./Testing/Ho1226_8894_0.wav\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         (None, 16000)             0         \n",
      "_________________________________________________________________\n",
      "reshape_1 (Reshape)          (None, 1, 16000)          0         \n",
      "_________________________________________________________________\n",
      "mel_stft (Melspectrogram)    (None, 80, 125, 1)        1091664   \n",
      "_________________________________________________________________\n",
      "normalization2d_1 (Normaliza (None, 80, 125, 1)        0         \n",
      "_________________________________________________________________\n",
      "permute_1 (Permute)          (None, 125, 80, 1)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 125, 80, 20)       120       \n",
      "_________________________________________________________________\n",
      "batch_normalization_1 (Batch (None, 125, 80, 20)       80        \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 62, 80, 20)        0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 99200)             0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                6348864   \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 32)                2080      \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 20)                660       \n",
      "=================================================================\n",
      "Total params: 7,443,468\n",
      "Trainable params: 6,351,764\n",
      "Non-trainable params: 1,091,704\n",
      "_________________________________________________________________\n",
      "(16000,)\n"
     ]
    }
   ],
   "source": [
    "if __name__ == \"__main__\":\n",
    "    #npary, sample_rate = _read_test_wav(\"Ho1226_8894_0.wav\")\n",
    "    #print(\"NumPy Arrary of testing file is {0} and sample rate is {1}\".format(npary, sample_rate))\n",
    "    #print(\"Loading the Model...........\")\n",
    "    #__load_model(_cnn_model_path)\n",
    "    main()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}