Liveliness_detection_1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Liveliness_detection_1.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOjmq5OtZB9BuIOkfGUFzkG",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ritikraj660/Speaker-recognition/blob/master/Liveliness_detection_1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c2A_kNLfBCln",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from pydrive.auth import GoogleAuth\n",
        "from pydrive.drive import GoogleDrive\n",
        "from google.colab import auth\n",
        "from oauth2client.client import GoogleCredentials"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UoZ_22nABJM9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "auth.authenticate_user()\n",
        "gauth = GoogleAuth()\n",
        "gauth.credentials = GoogleCredentials.get_application_default()\n",
        "drive = GoogleDrive(gauth)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cO9kbeMABLmR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "outputId": "a6ee5111-2c71-4b00-d8ad-96503d995aa0"
      },
      "source": [
        "\n",
        "link = 'https://drive.google.com/file/d/12fAst4usbQKCMaR6Io8OPNwna2Ye8S-o/view?usp=sharing'\n",
        "#fluff, id = link.split('=')\n",
        "id = '12fAst4usbQKCMaR6Io8OPNwna2Ye8S-o'\n",
        "print (id) "
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "12fAst4usbQKCMaR6Io8OPNwna2Ye8S-o\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qh5wm1NrBTsi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "downloaded = drive.CreateFile({'id':id}) \n",
        "downloaded.GetContentFile('liveliness.zip')  "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GJBcCgo3BbMK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 203
        },
        "outputId": "c9fcb377-fdca-4f47-9e49-e74a4aa4e12b"
      },
      "source": [
        "%%time\n",
        "!pip install patool\n",
        "import patoolib\n",
        "patoolib.extract_archive(\"liveliness.zip\")"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting patool\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/43/94/52243ddff508780dd2d8110964320ab4851134a55ab102285b46e740f76a/patool-1.12-py2.py3-none-any.whl (77kB)\n",
            "\r\u001b[K     |████▎                           | 10kB 21.8MB/s eta 0:00:01\r\u001b[K     |████████▌                       | 20kB 6.1MB/s eta 0:00:01\r\u001b[K     |████████████▊                   | 30kB 5.9MB/s eta 0:00:01\r\u001b[K     |█████████████████               | 40kB 6.9MB/s eta 0:00:01\r\u001b[K     |█████████████████████▏          | 51kB 6.5MB/s eta 0:00:01\r\u001b[K     |█████████████████████████▍      | 61kB 7.0MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▋  | 71kB 7.8MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 81kB 5.0MB/s \n",
            "\u001b[?25hInstalling collected packages: patool\n",
            "Successfully installed patool-1.12\n",
            "patool: Extracting liveliness.zip ...\n",
            "patool: running /usr/bin/7z x -o./Unpack_ybkndyr4 -- liveliness.zip\n",
            "patool: ... liveliness.zip extracted to `liveliness'.\n",
            "CPU times: user 29.2 ms, sys: 26.7 ms, total: 55.9 ms\n",
            "Wall time: 6.34 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OV4U8Z6GQOlH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 147
        },
        "outputId": "8eb4b472-2ab9-46ca-ad68-080eabd3a5f8"
      },
      "source": [
        "!pip install spafe\n",
        "from spafe.features.mfcc import mfcc"
      ],
      "execution_count": 161,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting spafe\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/a9/d9/d92807657a5125318639aa1636c0a76fb87222d48b5c290996ed9d5be1b6/spafe-0.1.2-py3-none-any.whl (77kB)\n",
            "\r\u001b[K     |████▎                           | 10kB 23.5MB/s eta 0:00:01\r\u001b[K     |████████▌                       | 20kB 6.3MB/s eta 0:00:01\r\u001b[K     |████████████▊                   | 30kB 8.8MB/s eta 0:00:01\r\u001b[K     |█████████████████               | 40kB 8.4MB/s eta 0:00:01\r\u001b[K     |█████████████████████▏          | 51kB 7.1MB/s eta 0:00:01\r\u001b[K     |█████████████████████████▍      | 61kB 8.1MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▋  | 71kB 8.4MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 81kB 5.4MB/s \n",
            "\u001b[?25hRequirement already satisfied: scipy>=1.3.1 in /usr/local/lib/python3.6/dist-packages (from spafe) (1.4.1)\n",
            "Requirement already satisfied: numpy>=1.17.2 in /usr/local/lib/python3.6/dist-packages (from spafe) (1.18.5)\n",
            "Installing collected packages: spafe\n",
            "Successfully installed spafe-0.1.2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ySJ3rJBcyyav",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install soundfile"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "idkDF-ldyA7j",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import librosa\n",
        "import soundfile\n",
        "import tqdm\n",
        "from scipy.io import wavfile\n",
        "import glob\n",
        "import os,pickle\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "import matplotlib.pyplot as plt\n",
        "import soundfile as sf\n",
        "import scipy\n",
        "from pathlib import Path\n",
        "import tensorflow.python.keras\n",
        "from tensorflow.python.keras.models import Sequential\n",
        "from tensorflow.python.keras.layers import Dense, Dropout, Flatten, BatchNormalization\n",
        "from keras.optimizers import SGD"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7P8v-4ClQE93",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from spafe.features.gfcc import gfcc\n",
        "\n",
        "def mfcc_(sig,fs):\n",
        "    mfccs  = mfcc(sig, fs=fs, num_ceps=20, pre_emph=1, pre_emph_coeff=0.95, \n",
        "              win_len=0.025, win_hop=0.0025, win_type='hamming', nfilts=26, \n",
        "              nfft=512, low_freq=14500, high_freq=15500, scale='constant', \n",
        "              dct_type=2, use_energy=True, lifter=22)\n",
        "    #print(fs)\n",
        "    return mfccs"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6RKRq9YXP7N8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def norm_(x,mean = False):\n",
        "    if mean == True:\n",
        "        return (x - np.mean(x)) / (np.max(x) - np.min(x)) #normalization                             \n",
        "    else:\n",
        "        return (x-np.mean(x)/np.std(x)) #standardization\n",
        "\n",
        "def zero_handling(x):\n",
        "    \"\"\"\n",
        "    handle the issue with zero values if they are exposed to become an argument\n",
        "    for any log function.\n",
        "\n",
        "    Returns:\n",
        "        vector with zeros substituted with epsilon values.\n",
        "    \"\"\"\n",
        "    return np.where(x == float(0), np.finfo(float).eps, x)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EYW3RaNHyl-1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def audioLengthInSecond(sr,m,t_s):#t_s will gives time in second[1 for 1sec, 2 for 2sec]\n",
        "    t = round(sr/512) # Number of frames in one second\n",
        "    totalFrame = round(t*t_s) # Number of frames in one second * time(second)\n",
        "    t1 = round(m.shape[0]/totalFrame) # total_audio_time/ t_s\n",
        "    #print(time_in_sec)\n",
        "    #print(\" Actual audio length in second :\",t1,\"\\n\",\"Number of frames in one second :\",t)\n",
        "    j=0\n",
        "    #print(totalFrame)\n",
        "    total = totalFrame*m.shape[1]\n",
        "    #print(m.shape)\n",
        "    result = np.zeros([1,total])\n",
        "    #print(total)\n",
        "    #print(result.shape)\n",
        "    for i in range(t1):\n",
        "            temp = m[j:j+totalFrame,:].reshape(1,-1)\n",
        "            diff = total-temp.shape[1]\n",
        "            if round(total*0.3)>diff: # if more 30% padding need done then we will not add that audio file\n",
        "                #print(diff)\n",
        "                if diff!=0:\n",
        "                    temp = np.concatenate((temp[0,:],temp[0,diff:diff+diff])).reshape(1,-1)\n",
        "                #print(\"temp\",temp.shape)\n",
        "                result = np.concatenate((result,temp))\n",
        "                j += totalFrame\n",
        "            else:\n",
        "                pass\n",
        "    return result[1:,:]\n",
        "#t_s=0.5\n",
        "#r = audioLengthInSecond(sr,contrast,t_s)\n",
        "#print(r.shape)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Um3KDwDwymsa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def label(user,total): # its creating lable for splited audio data\n",
        "    k_train = []\n",
        "    for i in range(total):\n",
        "        k_train.append(user)\n",
        "    y_train=np.array(k_train).T\n\n",
        "    return y_train"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rvIF-nkJSUdi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from pathlib import Path\n",
        "def s2c():\n",
        "    x,y=[],[]\n",
        "    t_s = 1\n",
        "    t = round(48000/512)\n",
        "    totalFrame = round(t*t_s)\n",
        "    total = totalFrame*20   \n",
        "    data = np.empty([1,total])\n",
        "    data_label = np.asarray([])\n",
        "    count =0  # E:\\dataset\\LibriSpeech\\dev-clean\n",
        "    for file in Path('liveliness').glob('*.wav'):\n",
        "        file_name=os.path.basename(file)\n",
        "        user = file_name[0]\n",
        "        #print(f)\n",
        "      \n",
        "        with soundfile.SoundFile(file) as sound_file:\n",
        "                signal = sound_file.read(dtype=\"float32\")\n",
        "                sr=sound_file.samplerate\n",
        "        signal = norm_(signal,mean = False)\n",
        "        signal = zero_handling(signal)\n",
        " \n",
        "        feature = mfcc_(signal,sr)\n",
        "              #print(feature.shape[0])\n",
        "        re_data = audioLengthInSecond(sr,feature,t_s) # return (3,400)\n",
        "        #print(\"shape of feature return \",re_data.shape)\n",
        "        data=np.concatenate((data,re_data),axis=0)\n",
        "        #print(\"Updated data length\",data.shape)\n",
        "        re_label = label(user,re_data.shape[0])\n",
        "        data_label = np.concatenate((data_label,re_label),axis=0)\n",
        "    return data[1:,:],data_label"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tPX2mC1ty2mq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def load_data1():\n",
        "    #x,y=[],[]   #C:\\Users\\RITIK_RAJ\\Music\\liveliness\\*.wav\n",
        "    #for file in glob.glob(\"C:\\\\Users\\\\RITIK_RAJ\\\\Music\\\\liveliness//*\\\\*.wav\"):   \n",
        "    print(\"Entered into pre-processing step\")\n",
        "    data = np.empty([1,329])  #658, #329\n",
        "    data_label = np.asarray([])\n",
        "    for file in Path('liveliness').glob('*.wav'):\n",
        "        file_name=os.path.basename(file)\n",
        "        #print(\"user_name\", file_name[0])\n",
        "        user = file_name[0]\n",
        "        with soundfile.SoundFile(file) as sound_file:\n",
        "            X = sound_file.read(dtype=\"float32\")\n",
        "            sample_rate=sound_file.samplerate\n",
        "        #print(\"Length of Audio = \",X.shape,\" Sampling Rate = \",sample_rate)\n",
        "        X = norm_(X,mean = False)\n",
        "        X = zero_handling(X)\n",
        "        psi = scipy.signal.butter(4, [14500, 15500], btype='bandpass', output='sos', fs=sample_rate)\n",
        "        # filter audio\n",
        "        y_filtered = scipy.signal.sosfilt(psi, X)\n",
        "        #print(\"Butter_worth_filter\",y_filtered.shape)\n",
        "        S = np.abs(librosa.stft(y_filtered, hop_length= 256, win_length =1024 ))\n",
        "        #print(\"STFT shape = \",S.shape)\n",
        "        contrast = librosa.feature.spectral_contrast(S=S, sr=sample_rate).T\n",
        "        #print(\"Feature extraction \",contrast.shape)\n",
        "        r = audioLengthInSecond(sample_rate,contrast,t_s=0.5)\n",
        "        #print(r.shape)\n",
        "        data=np.concatenate((data,r),axis=0)\n",
        "        re_label = label(user,r.shape[0])\n",
        "        data_label = np.concatenate((data_label,re_label),axis=0)\n",
        "    print(\"Exiting Pre-processing step\")    \n",
        "    return data[1:,:],data_label"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LwOkUuliy5GU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#x,y = load_data()\n",
        "#print(x.shape,y.shape)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K53ADQdry82j",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def data_split():\n",
        "  #X  = x.reshape(x.shape[0],x.shape[1],x.shape[2],1)\n",
        "  x,y = s2c()#load_data()\n",
        "  u = np.unique(y)\n",
        "  num_cla = len(u)\n",
        "  #print(u)\n",
        "  d = dict()\n",
        "  for enu,key in enumerate(u,0):\n",
        "      d[key]=enu\n",
        "  #print(d)\n",
        "  for i in range(len(y)):\n",
        "      y[i] = d[y[i]]  \n",
        "  print(x.shape,y.shape)\n",
        "  print(\"Shape of Input data \",x.shape)\n",
        "  import keras\n",
        "\n",
        "  y_binary = keras.utils.to_categorical(y, num_classes = num_cla)\n",
        "  #y_binary = np.asarray(list(map(int,y)))\n",
        "  #print(y_binary)\n",
        "\n",
        "  x_train, x_test, y_train, y_test = train_test_split(x,y_binary ,test_size=0.20, shuffle= True)\n",
        "\n",
        "  print(\"Training shape: {0}, Training label shape : {1}\".format(x_train.shape,y_train.shape))\n",
        "  print(\"Training shape: {0}, Training label shape : {1}\".format(x_test.shape,y_test.shape))\n",
        "  return  x_train, x_test, y_train, y_test"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "okgD0oTfzalI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import keras\n",
        "from matplotlib import pyplot as plt\n",
        "def accuracy_plot(history):\n",
        "    plt.figure(figsize=(15,8))\n",
        "    plt.plot(history.history['accuracy'])\n",
        "    plt.plot(history.history['val_accuracy'])\n",
        "    plt.title('model accuracy')\n",
        "    plt.ylabel('accuracy')\n",
        "    plt.xlabel('epoch')\n",
        "    plt.legend(['train', 'val'], loc='upper left')\n",
        "    plt.grid()\n",
        "    plt.show()\n",
        "\n",
        "def loss_plot(history):\n",
        "    plt.figure(figsize=(15,8))\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', 'val'], loc='upper left')\n",
        "    plt.grid()\n",
        "    plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c2E5DuSFzdbx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.optimizers import SGD\n",
        "from tensorflow.python.keras.callbacks import ModelCheckpoint\n",
        "def model_(x,outputsize):\n",
        "  model = Sequential()\n",
        "  model.add(Dense(64, input_shape=(x.shape[1],), activation='relu'))\n",
        "  model.add(Dropout(0.5))\n",
        "  model.add(Dense(64, activation='relu'))\n",
        "  model.add(Dropout(0.5))\n",
        "  model.add(Dense(64, activation='relu'))\n",
        "  model.add(Dropout(0.5))\n",
        "  model.add(Dense(outputsize, activation='softmax'))\n",
        "  #print(model.summary())\n",
        "  opt = SGD(lr=0.001, momentum=0.9)                        #optimizer=keras.optimizers.Adam()\n",
        "  #opt =Adam(learning_rate=0.001)\n",
        "  model.compile(loss=keras.losses.categorical_crossentropy, optimizer=opt, metrics=['accuracy'])\n",
        "  #tensorflow.keras.utils.plot_model(model, to_file='model.png', show_shapes=False, show_layer_names=True,\n",
        "  #                         rankdir='TB', expand_nested=False, dpi=96)\n",
        "  print(model.summary())\n",
        "  return model\n",
        "  "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oa--_AGk0MU8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas.util.testing as tm\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "def conf(y_test,yhat):\n",
        "  y_test = np.argmax(y_test, axis=1)\n",
        "  total_user = ['Live','Replay']\n",
        "  #total_user = [0,1]\n",
        "  conf_mat = confusion_matrix(y_test, yhat)\n",
        "  fig, ax = plt.subplots(figsize=(10,10))\n",
        "  sns.heatmap(conf_mat, cmap=\"YlGnBu\",annot=True, fmt='d',\n",
        "              xticklabels=total_user, yticklabels=total_user)\n",
        "  plt.ylabel('Actual')\n",
        "  plt.xlabel('Predicted')\n",
        "  plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rZjmuYAgBAaH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lY2Tqsn70ek3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def training(model,x_train, x_test, y_train, y_test,epoch = 300,batch_sizes = 16):\n",
        "  # Saving the model that performed the best on the validation set\n",
        "  checkpoint = ModelCheckpoint(filepath='Model.v_2.hdf5', save_best_only=True, verbose=0)\n",
        "\n",
        "  #validation_split=0.2\n",
        "\n",
        "  #history = model.fit(x_train, y_train, batch_size = batch_sizes, epochs=epoch, \n",
        "  #                    validation_data=(x_test, y_test), verbose=0, callbacks=[checkpoint])\n",
        "  history = model.fit(x_train, y_train, batch_size = batch_sizes, epochs=epoch, \n",
        "                      validation_split=0.2, verbose=0, callbacks=[checkpoint])\n",
        "  #Loading the model that performed the best on the validation set\n",
        "  model.load_weights('Model.v_2.hdf5')\n",
        "\n",
        "  # Testing the model on the Test data\n",
        "  (loss, accuracy) = model.evaluate(x_test, y_test, batch_size=8, verbose=1)\n",
        "\n",
        "  yhat = np.argmax(model.predict(x_test), axis=-1)\n",
        "  conf(y_test,yhat)\n",
        "  accuracy_plot(history)\n",
        "  loss_plot(history)\n",
        "  print('Accuracy on test data: {:.2f}%'.format(accuracy * 100))\n",
        "  print('Loss on test data: {:.2f}%'.format(loss*100))\n",
        "  return [accuracy*100,loss*100]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X3dMrevE0nQr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "9f94b2ab-7e54-4d57-c016-540328a03a1e"
      },
      "source": [
        "%%time\n",
        "class speaker:\n",
        "\n",
        "  def __init__(self,epoch,batch_sizes):\n",
        "    self.epoch = epoch\n",
        "    self.batch_sizes = batch_sizes\n",
        "    self.run()\n",
        "\n",
        "\n",
        "  def run(self):\n",
        "    x_train, x_test, y_train, y_test = data_split()\n",
        "    print(\"Entering making model\")\n",
        "    model = model_(x_train,2)\n",
        "    print(\"Enteres training phase\")\n",
        "    #l.append(training(model, x_train, x_test, y_train, y_test,self.epoch, self.batch_sizes))\n",
        "    training(model, x_train, x_test, y_train, y_test,self.epoch, self.batch_sizes)\n",
        "    #print(\"Accuracy of model {}, Loss of model {}\".format(l[0],l[1]))\n"
      ],
      "execution_count": 302,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CPU times: user 35 µs, sys: 3 µs, total: 38 µs\n",
            "Wall time: 42.7 µs\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j8-EPHAJ0rdH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "c109ee1e-5435-44e8-c0e4-f5f386802bde"
      },
      "source": [
        "%%time\n",
        "speaker(epoch = 100, batch_sizes = 16)"
      ],
      "execution_count": 304,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(5957, 1880) (5957,)\n",
            "Shape of Input data  (5957, 1880)\n",
            "Training shape: (4765, 1880), Training label shape : (4765, 2)\n",
            "Training shape: (1192, 1880), Training label shape : (1192, 2)\n",
            "Entering making model\n",
            "Model: \"sequential_26\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_104 (Dense)            (None, 64)                120384    \n",
            "_________________________________________________________________\n",
            "dropout_54 (Dropout)         (None, 64)                0         \n",
            "_________________________________________________________________\n",
            "dense_105 (Dense)            (None, 64)                4160      \n",
            "_________________________________________________________________\n",
            "dropout_55 (Dropout)         (None, 64)                0         \n",
            "_________________________________________________________________\n",
            "dense_106 (Dense)            (None, 64)                4160      \n",
            "_________________________________________________________________\n",
            "dropout_56 (Dropout)         (None, 64)                0         \n",
            "_________________________________________________________________\n",
            "dense_107 (Dense)            (None, 2)                 130       \n",
            "=================================================================\n",
            "Total params: 128,834\n",
            "Trainable params: 128,834\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Enteres training phase\n",
            "149/149 [==============================] - 0s 1ms/step - loss: 0.4959 - accuracy: 0.7659\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Accuracy on test data: 76.59%\n",
            "Loss on test data: 49.59%\n",
            "CPU times: user 1min 21s, sys: 7.1 s, total: 1min 28s\n",
            "Wall time: 1min 3s\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<__main__.speaker at 0x7fdc4f0db438>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 304
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KYl4L3cP1m-9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}