Inference-soft.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.io.wavfile import read as read_wav\n",
    "from IPython.display import Audio\n",
    "import matplotlib.pyplot as plt\n",
    "from collections import Counter\n",
    "import tensorflow as tf\n",
    "import librosa as lr\n",
    "import seaborn as sn\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy\n",
    "import pydub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "genres = {\n",
    "    'disco': 0,\n",
    "    'country': 1,\n",
    "    'jazz': 2,\n",
    "    'classical': 3,\n",
    "    'blues': 4,\n",
    "    'hiphop': 5,\n",
    "    'metal': 6,\n",
    "    'rock': 7,\n",
    "    'pop': 8,\n",
    "    'reggae': 9\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "hparams = {\n",
    "    'samplerate': 22050,\n",
    "    'seq_length': 256,\n",
    "    'hop_length': 512,\n",
    "    'fft_window': 2048\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_features(filepath, hparams):\n",
    "    \n",
    "    audio_data_array = lr.load(filepath, sr=hparams['samplerate'], mono=True)\n",
    "    \n",
    "    # normalized = lr.util.normalize(audio_data_array[0])\n",
    "    \n",
    "    features = [\n",
    "        \n",
    "        # MFCC Features:\n",
    "        lr.feature.mfcc(\n",
    "            y          = audio_data_array[0],\n",
    "            sr         = hparams['samplerate'],\n",
    "            hop_length = hparams['hop_length'],\n",
    "            n_fft      = hparams['fft_window']\n",
    "        ),\n",
    "\n",
    "        # Chroma STFT:\n",
    "        lr.feature.chroma_stft(\n",
    "            y          = audio_data_array[0],\n",
    "            sr         = hparams['samplerate'],\n",
    "            hop_length = hparams['hop_length'],\n",
    "            n_fft      = hparams['fft_window']\n",
    "        ),\n",
    "        \n",
    "        # Spectral Centroid:\n",
    "        lr.feature.spectral_centroid(\n",
    "            y          = audio_data_array[0],\n",
    "            sr         = hparams['samplerate'],\n",
    "            hop_length = hparams['hop_length'],\n",
    "            n_fft      = hparams['fft_window']\n",
    "        ),\n",
    "        \n",
    "        # Spectral Contrast:\n",
    "        lr.feature.spectral_contrast(\n",
    "            y          = audio_data_array[0],\n",
    "            sr         = hparams['samplerate'],\n",
    "            hop_length = hparams['hop_length'],\n",
    "            n_fft      = hparams['fft_window']\n",
    "        )\n",
    "    ]\n",
    "    \n",
    "    return np.vstack(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "lstm_1_model = tf.keras.models.load_model('LSTM-1/models/model')\n",
    "lstm_2a_model = tf.keras.models.load_model('LSTM-2a/models/model')\n",
    "lstm_2b_model = tf.keras.models.load_model('LSTM-2b/models/model')\n",
    "lstm_3a_model = tf.keras.models.load_model('LSTM-3a/models/model')\n",
    "lstm_3b_model = tf.keras.models.load_model('LSTM-3b/models/model')\n",
    "lstm_3c_model = tf.keras.models.load_model('LSTM-3c/models/model')\n",
    "lstm_3d_model = tf.keras.models.load_model('LSTM-3d/models/model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def lookup(arr):\n",
    "    if arr == [0, 0, 0]:\n",
    "        return 'hiphop'\n",
    "    elif arr == [0, 0, 1]:\n",
    "        return 'metal'\n",
    "    elif arr == [0, 0, 2]:\n",
    "        return 'rock'\n",
    "    elif arr == [0, 1, 0]:\n",
    "        return 'pop'\n",
    "    elif arr == [0, 1, 1]:\n",
    "        return 'reggae'\n",
    "    elif arr == [1, 0, 0]:\n",
    "        return 'disco'\n",
    "    elif arr == [1, 0, 1]:\n",
    "        return 'country'\n",
    "    elif arr == [1, 1, 0]:\n",
    "        return 'jazz'\n",
    "    elif arr == [1, 1, 1]:\n",
    "        return 'classical'\n",
    "    else:\n",
    "        return 'blues'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(features):\n",
    "    \n",
    "    result = []\n",
    "    \n",
    "    lstm_1_res = np.argmax(lstm_1_model.predict(features))\n",
    "    # print('LSTM-1:', lstm_1_res)\n",
    "    \n",
    "    result += [lstm_1_res]\n",
    "    \n",
    "    # strong classes:\n",
    "    if lstm_1_res == 0:\n",
    "        \n",
    "        lstm_2a_res = np.argmax(lstm_2a_model.predict(features))\n",
    "        # print('LSTM-2a:', lstm_2a_res)\n",
    "        \n",
    "        result += [lstm_2a_res]\n",
    "        \n",
    "        # substrong-1:\n",
    "        if lstm_2a_res == 0:\n",
    "            \n",
    "            lstm_3a_res = np.argmax(lstm_3a_model.predict(features))\n",
    "            # print('LSTM-3a:', lstm_3a_res)\n",
    "            \n",
    "            result += [lstm_3a_res]\n",
    "            \n",
    "        # substrong-2:\n",
    "        else:\n",
    "            \n",
    "            lstm_3b_res = np.argmax(lstm_3b_model.predict(features))\n",
    "            # print('LSTM-3b:', lstm_3b_res)\n",
    "            \n",
    "            result += [lstm_3b_res]\n",
    "            \n",
    "    # mild classes:\n",
    "    else:\n",
    "        \n",
    "        lstm_2b_res = np.argmax(lstm_2b_model.predict(features))\n",
    "        # print('LSTM-2b:', lstm_2b_res)\n",
    "        \n",
    "        result += [lstm_2b_res]\n",
    "        \n",
    "        \n",
    "        # submild-1:\n",
    "        if lstm_2b_res == 0:\n",
    "            \n",
    "            lstm_3c_res = np.argmax(lstm_3c_model.predict(features))\n",
    "            # print('LSTM-3c:', lstm_3c_res)\n",
    "            \n",
    "            result += [lstm_3c_res]\n",
    "            \n",
    "        # submild-2:\n",
    "        else:\n",
    "            \n",
    "            lstm_3d_res = np.argmax(lstm_3d_model.predict(features))\n",
    "            # print('LSTM-3d:', lstm_3d_res)\n",
    "            \n",
    "            result += [lstm_3d_res]\n",
    "            \n",
    "    return lookup(list(result))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Accuracy on GTZAN Validation and Test Set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "processed = np.load('./prepared/processed.npy', allow_pickle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Top-1 Accuracy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_1_confusion_mat = np.zeros(shape=(10, 10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "ground_truth_labels = []\n",
    "\n",
    "for i in range(0, 5000, 5):\n",
    "    \n",
    "    if (i % 500) >= 5 * 90:\n",
    "        predicted = []\n",
    "        for j in range(0, 5):\n",
    "            if ((i+j) % 500) >= 5 * 90:\n",
    "                predicted += [predict(processed[i+j]['features'].reshape(1, 256, 40))]\n",
    "        \n",
    "        original = processed[i]['ground_truth']\n",
    "        \n",
    "        top_1 = Counter(predicted).most_common(n=1)\n",
    "        \n",
    "        if original == top_1[0][0]:\n",
    "            top_1_confusion_mat[genres[original]][genres[original]] += 1\n",
    "        else:\n",
    "            top_1_confusion_mat[genres[original]][genres[top_1[0][0]]] += 1\n",
    "        \n",
    "        ground_truth_labels += [original]\n",
    "        \n",
    "        # print('=========================================')\n",
    "        # print('Original: {}'.format(original))\n",
    "        # print('Predicted: {}'.format(top_1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_1_frame = pd.DataFrame(\n",
    "    data=top_1_confusion_mat,\n",
    "    index=list(genres.keys()),\n",
    "    columns=list(genres.keys())\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 6))\n",
    "\n",
    "sn.set(font_scale=1.4) # for label size\n",
    "sn.heatmap(top_1_frame, annot=True, annot_kws={\"size\": 10})\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top-1 Accuracy: 63.0 %\n"
     ]
    }
   ],
   "source": [
    "top_1_accuracy = 100 * (np.sum(np.diag(top_1_confusion_mat)) / np.sum(top_1_confusion_mat))\n",
    "print('Top-1 Accuracy: {} %'.format(top_1_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Top-2 Accuracy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_2_confusion_mat = np.zeros(shape=(10, 10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "for i in range(0, 5000, 5):\n",
    "    \n",
    "    if (i % 500) >= 5 * 90:\n",
    "        predicted = []\n",
    "        for j in range(0, 5):\n",
    "            if ((i+j) % 500) >= 5 * 90:\n",
    "                predicted += [predict(processed[i+j]['features'].reshape(1, 256, 40))]\n",
    "        \n",
    "        original = processed[i]['ground_truth']\n",
    "        \n",
    "        top_2 = Counter(predicted).most_common(n=2)\n",
    "        \n",
    "        if original in [top_2[i][0] for i in range(len(top_2))]:\n",
    "            top_2_confusion_mat[genres[original]][genres[original]] += 1\n",
    "        else:\n",
    "            for i in range(len(top_2)):\n",
    "                top_2_confusion_mat[genres[original]][genres[top_2[i][0]]] += (1 / len(top_2))\n",
    "        \n",
    "        # print('=========================================')\n",
    "        # print('Original: {}'.format(original))\n",
    "        # print('Predicted: {}'.format(top_2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_2_frame = pd.DataFrame(\n",
    "    data=top_2_confusion_mat,\n",
    "    index=list(genres.keys()),\n",
    "    columns=list(genres.keys())\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 6))\n",
    "\n",
    "sn.set(font_scale=1.4) # for label size\n",
    "sn.heatmap(top_2_frame, annot=True, annot_kws={\"size\": 10})\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top-2 Accuracy: 80.0 %\n"
     ]
    }
   ],
   "source": [
    "top_2_accuracy = 100 * (np.sum(np.diag(top_2_confusion_mat)) / np.sum(top_2_confusion_mat))\n",
    "print('Top-2 Accuracy: {} %'.format(top_2_accuracy))"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}