{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sVNOuHQQjsrp"
   },
   "source": [
    "# PyTorch Example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9ZTzCtc5kYVg"
   },
   "source": [
    "## Install Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T14:00:15.701867Z",
     "start_time": "2020-12-15T14:00:09.512876Z"
    },
    "cellView": "form",
    "id": "rllMjjsekbjt"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://github.com/snakers4/silero-vad/archive/master.zip\" to /home/keras/.cache/torch/hub/master.zip\n",
      "/opt/conda/lib/python3.8/site-packages/torchaudio/backend/utils.py:53: UserWarning: \"sox\" backend is being deprecated. The default backend will be changed to \"sox_io\" backend in 0.8.0 and \"sox\" backend will be removed in 0.9.0. Please migrate to \"sox_io\" backend. Please refer to https://github.com/pytorch/audio/issues/903 for the detail.\n",
      "  warnings.warn(\n",
      "/opt/conda/lib/python3.8/site-packages/torchaudio/backend/utils.py:63: UserWarning: The interface of \"soundfile\" backend is planned to change in 0.8.0 to match that of \"sox_io\" backend and the current interface will be removed in 0.9.0. To use the new interface, do `torchaudio.USE_SOUNDFILE_LEGACY_INTERFACE = False` before setting the backend to \"soundfile\". Please refer to https://github.com/pytorch/audio/issues/903 for the detail.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "#@title Install and Import Dependencies\n",
    "\n",
    "# this assumes that you have a relevant version of PyTorch installed\n",
    "#!pip install -q torchaudio soundfile\n",
    "\n",
    "import glob\n",
    "import torch\n",
    "torch.set_num_threads(1)\n",
    "\n",
    "from IPython.display import Audio\n",
    "\n",
    "\n",
    "model, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
    "                              model='silero_vad',\n",
    "                              force_reload=True)\n",
    "\n",
    "(get_speech_ts,\n",
    " save_audio,\n",
    " read_audio,\n",
    " state_generator,\n",
    " single_audio_stream,\n",
    " collect_speeches) = utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T14:03:38.006309Z",
     "start_time": "2020-12-15T14:03:38.002613Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/home/keras/.cache/torch/hub'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.hub.get_dir()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fXbbaUO3jsrw"
   },
   "source": [
    "## Full audio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:56.879818Z",
     "start_time": "2020-12-15T13:09:56.864765Z"
    },
    "id": "aI_eydBPjsrx"
   },
   "outputs": [],
   "source": [
    "wav = read_audio('files/en.wav')\n",
    "# get speech timestamps from full audio file\n",
    "speech_timestamps = get_speech_ts(wav, model,\n",
    "                                  num_steps=4)\n",
    "print(speech_timestamps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:58.941063Z",
     "start_time": "2020-12-15T13:09:58.887006Z"
    },
    "id": "OuEobLchjsry"
   },
   "outputs": [],
   "source": [
    "# merge all speech chunks to one audio\n",
    "save_audio('only_speech.wav',\n",
    "           collect_speeches(speech_timestamps, wav), 16000) \n",
    "Audio('only_speech.wav')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iDKQbVr8jsry"
   },
   "source": [
    "## Single Audio Stream"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:59.199321Z",
     "start_time": "2020-12-15T13:09:59.196823Z"
    },
    "id": "q-lql_2Wjsry"
   },
   "outputs": [],
   "source": [
    "wav = 'files/en.wav'\n",
    "\n",
    "for batch in single_audio_stream(model, wav):\n",
    "    if batch:\n",
    "        print(batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KBDVybJCjsrz"
   },
   "source": [
    "## Multiple Audio Streams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:10:03.590358Z",
     "start_time": "2020-12-15T13:10:03.587071Z"
    },
    "id": "BK4tGfWgjsrz"
   },
   "outputs": [],
   "source": [
    "audios_for_stream = glob.glob('files/*.wav')\n",
    "len(audios_for_stream) # total 4 audios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:10:15.762491Z",
     "start_time": "2020-12-15T13:10:03.591388Z"
    },
    "id": "v1l8sam1jsrz"
   },
   "outputs": [],
   "source": [
    "for batch in state_generator(model, audios_for_stream, audios_in_stream=2): # 2 audio stream\n",
    "    if batch:\n",
    "        print(batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "57avIBd6jsrz"
   },
   "source": [
    "# ONNX Example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bL4kn4KJrlyL"
   },
   "source": [
    "## Install Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Q4QIfSpprnkI"
   },
   "outputs": [],
   "source": [
    "#@title Install and Import Dependencies\n",
    "\n",
    "# this assumes that you have a relevant version of PyTorch installed\n",
    "!pip install -q torchaudio soundfile onnxruntime\n",
    "\n",
    "import glob\n",
    "import onnxruntime\n",
    "\n",
    "from IPython.display import Audio\n",
    "\n",
    "_, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
    "                              model='silero_vad',\n",
    "                              force_reload=True)\n",
    "\n",
    "(get_speech_ts,\n",
    " save_audio,\n",
    " read_audio,\n",
    " state_generator,\n",
    " single_audio_stream,\n",
    " collect_speeches) = utils\n",
    "\n",
    " def init_onnx_model(model_path: str):\n",
    "    return onnxruntime.InferenceSession(model_path)\n",
    "\n",
    "def validate_onnx(model, inputs):\n",
    "    with torch.no_grad():\n",
    "        ort_inputs = {'input': inputs.cpu().numpy()}\n",
    "        outs = model.run(None, ort_inputs)\n",
    "        outs = [torch.Tensor(x) for x in outs]\n",
    "    return outs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5JHErdB7jsr0"
   },
   "source": [
    "## Full audio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:06.643812Z",
     "start_time": "2020-12-15T13:09:06.473386Z"
    },
    "id": "krnGoA6Kjsr0"
   },
   "outputs": [],
   "source": [
    "model = init_onnx_model('files/model.onnx')\n",
    "wav = read_audio('files/en.wav')\n",
    "\n",
    "# get speech timestamps from full audio file\n",
    "speech_timestamps = get_speech_ts(wav, model, num_steps=4, run_function=validate_onnx) \n",
    "print(speech_timestamps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:08.862421Z",
     "start_time": "2020-12-15T13:09:08.820014Z"
    },
    "id": "B176Lzfnjsr1"
   },
   "outputs": [],
   "source": [
    "# merge all speech chunks to one audio\n",
    "save_audio('only_speech.wav', collect_speeches(speech_timestamps, wav), 16000)\n",
    "Audio('only_speech.wav')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Rio9W50gjsr1"
   },
   "source": [
    "## Single audio stream"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:09.606031Z",
     "start_time": "2020-12-15T13:09:09.504239Z"
    },
    "id": "IPkl8Yy1jsr1"
   },
   "outputs": [],
   "source": [
    "model = init_onnx_model('files/model.onnx')\n",
    "wav = 'files/en.wav'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:11.453171Z",
     "start_time": "2020-12-15T13:09:09.633435Z"
    },
    "id": "NC6Jim0hjsr1"
   },
   "outputs": [],
   "source": [
    "for batch in single_audio_stream(model, wav, run_function=validate_onnx):\n",
    "    if batch:\n",
    "        print(batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WNZ42u0ajsr1"
   },
   "source": [
    "## Multiple audio stream"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:11.540423Z",
     "start_time": "2020-12-15T13:09:11.455706Z"
    },
    "id": "XjhGQGppjsr1"
   },
   "outputs": [],
   "source": [
    "model = init_onnx_model('files/model.onnx')\n",
    "audios_for_stream = glob.glob('files/*.wav')\n",
    "print(len(audios_for_stream)) # total 4 audios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-12-15T13:09:19.565434Z",
     "start_time": "2020-12-15T13:09:11.552097Z"
    },
    "id": "QI7-arlqjsr2"
   },
   "outputs": [],
   "source": [
    "for batch in state_generator(model, audios_for_stream, audios_in_stream=2, run_function=validate_onnx): # 2 audio stream\n",
    "    if batch:\n",
    "        print(batch)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "silero-vad.ipynb",
   "provenance": []
  },
  "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": "3.8.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}