deepgeek/silero-vad
2
0
Fork 0
mirror of https://github.com/snakers4/silero-vad.git synced 2026-02-04 17:39:22 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
c762bb5b529d24102a4dee94380f853142c08bea
silero-vad/silero-vad.ipynb
adamnsandle c762bb5b52 add adaptive examples
2021-04-15 14:01:05 +00:00

1115 lines
29 KiB
Plaintext
Executable File
Raw Blame History

{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"id": "sVNOuHQQjsrp"
},
"source": [
"# PyTorch Examples"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## VAD"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:35:43.397137Z",
"start_time": "2020-12-30T17:33:10.962078Z"
},
"hidden": true
},
"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\n",
"\n",
"import glob\n",
"import torch\n",
"torch.set_num_threads(1)\n",
"\n",
"from IPython.display import Audio\n",
"from pprint import pprint\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",
" get_speech_ts_adaptive,\n",
" save_audio,\n",
" read_audio,\n",
" state_generator,\n",
" single_audio_stream,\n",
" collect_chunks) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "fXbbaUO3jsrw"
},
"source": [
"### Full Audio"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:35:44.362860Z",
"start_time": "2020-12-30T17:35:43.398441Z"
},
"hidden": true,
"id": "aI_eydBPjsrx"
},
"outputs": [],
"source": [
"wav = read_audio(f'{files_dir}/en.wav')\n",
"# get speech timestamps from full audio file\n",
"speech_timestamps = get_speech_ts(wav, model,\n",
" num_steps=4)\n",
"pprint(speech_timestamps)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:35:44.419280Z",
"start_time": "2020-12-30T17:35:44.364175Z"
},
"hidden": true,
"id": "OuEobLchjsry"
},
"outputs": [],
"source": [
"# merge all speech chunks to one audio\n",
"save_audio('only_speech.wav',\n",
" collect_chunks(speech_timestamps, wav), 16000) \n",
"Audio('only_speech.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"wav = read_audio(f'{files_dir}/en.wav')\n",
"# get speech timestamps from full audio file\n",
"speech_timestamps = get_speech_ts_adaptive(wav, model, step=500, num_samples_per_window=4000)\n",
"pprint(speech_timestamps)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"# merge all speech chunks to one audio\n",
"save_audio('only_speech.wav',\n",
" collect_chunks(speech_timestamps, wav), 16000) \n",
"Audio('only_speech.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "iDKQbVr8jsry"
},
"source": [
"### Single Audio Stream"
]
},
{
"cell_type": "markdown",
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:29:04.224833Z",
"start_time": "2021-04-15T13:29:04.220588Z"
},
"hidden": true
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:59.199321Z",
"start_time": "2020-12-15T13:09:59.196823Z"
},
"hidden": true,
"id": "q-lql_2Wjsry"
},
"outputs": [],
"source": [
"wav = f'{files_dir}/en.wav'\n",
"\n",
"for batch in single_audio_stream(model, wav):\n",
" if batch:\n",
" print(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"wav = f'{files_dir}/en.wav'\n",
"\n",
"for batch in single_audio_stream(model, wav, iterator_type='adaptive'):\n",
" if batch:\n",
" print(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"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"
},
"hidden": true,
"id": "BK4tGfWgjsrz"
},
"outputs": [],
"source": [
"audios_for_stream = glob.glob(f'{files_dir}/*.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"
},
"hidden": true,
"id": "v1l8sam1jsrz"
},
"outputs": [],
"source": [
"for batch in state_generator(model, audios_for_stream, audios_in_stream=2): # 2 audio stream\n",
" if batch:\n",
" pprint(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## Number detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"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\n",
"\n",
"import glob\n",
"import torch\n",
"torch.set_num_threads(1)\n",
"\n",
"from IPython.display import Audio\n",
"from pprint import pprint\n",
"\n",
"model, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
" model='silero_number_detector',\n",
" force_reload=True)\n",
"\n",
"(get_number_ts,\n",
" save_audio,\n",
" read_audio,\n",
" collect_chunks,\n",
" drop_chunks) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"### Full audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"wav = read_audio(f'{files_dir}/en_num.wav')\n",
"# get number timestamps from full audio file\n",
"number_timestamps = get_number_ts(wav, model)\n",
"pprint(number_timestamps)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"sample_rate = 16000\n",
"# convert ms in timestamps to samples\n",
"for timestamp in number_timestamps:\n",
" timestamp['start'] = int(timestamp['start'] * sample_rate / 1000)\n",
" timestamp['end'] = int(timestamp['end'] * sample_rate / 1000)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"# merge all number chunks to one audio\n",
"save_audio('only_numbers.wav',\n",
" collect_chunks(number_timestamps, wav), sample_rate) \n",
"Audio('only_numbers.wav')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"# drop all number chunks from audio\n",
"save_audio('no_numbers.wav',\n",
" drop_chunks(number_timestamps, wav), sample_rate) \n",
"Audio('no_numbers.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## Language detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"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\n",
"\n",
"import glob\n",
"import torch\n",
"torch.set_num_threads(1)\n",
"\n",
"from IPython.display import Audio\n",
"from pprint import pprint\n",
"\n",
"model, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
" model='silero_lang_detector',\n",
" force_reload=True)\n",
"\n",
"(get_language,\n",
" read_audio) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"### Full audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"wav = read_audio(f'{files_dir}/en.wav')\n",
"lang = get_language(wav, model)\n",
"print(lang)"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"id": "57avIBd6jsrz"
},
"source": [
"# ONNX Example"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## VAD"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "bL4kn4KJrlyL"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:30:22.938755Z",
"start_time": "2021-04-15T13:30:20.970574Z"
},
"cellView": "form",
"hidden": true,
"id": "Q4QIfSpprnkI"
},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'torch' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m\u001b[0m",
"\u001b[0;31mNameError\u001b[0mTraceback (most recent call last)",
"\u001b[0;32m<ipython-input-3-ca9e92528117>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 10\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mIPython\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdisplay\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mAudio\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 11\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m _, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n\u001b[0m\u001b[1;32m 13\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'silero_vad'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 14\u001b[0m force_reload=True)\n",
"\u001b[0;31mNameError\u001b[0m: name 'torch' is not defined"
]
}
],
"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",
"from pprint import pprint\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",
" get_speech_ts_adaptive,\n",
" save_audio,\n",
" read_audio,\n",
" state_generator,\n",
" single_audio_stream,\n",
" collect_speeches) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'\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[0]"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "5JHErdB7jsr0"
},
"source": [
"### Full Audio"
]
},
{
"cell_type": "markdown",
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:34:22.554010Z",
"start_time": "2021-04-15T13:34:22.550308Z"
},
"hidden": true
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:30:14.475412Z",
"start_time": "2021-04-15T13:30:14.427933Z"
},
"hidden": true,
"id": "krnGoA6Kjsr0"
},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'init_onnx_model' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m\u001b[0m",
"\u001b[0;31mNameError\u001b[0mTraceback (most recent call last)",
"\u001b[0;32m<ipython-input-2-65cde4c4cba8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0minit_onnx_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf'{files_dir}/model.onnx'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mwav\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mread_audio\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf'{files_dir}/en.wav'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0;31m# get speech timestamps from full audio file\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0mspeech_timestamps\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_speech_ts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwav\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_steps\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrun_function\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidate_onnx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mNameError\u001b[0m: name 'init_onnx_model' is not defined"
]
}
],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"wav = read_audio(f'{files_dir}/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",
"pprint(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"
},
"hidden": true,
"id": "B176Lzfnjsr1"
},
"outputs": [],
"source": [
"# merge all speech chunks to one audio\n",
"save_audio('only_speech.wav', collect_chunks(speech_timestamps, wav), 16000)\n",
"Audio('only_speech.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"wav = read_audio(f'{files_dir}/en.wav')\n",
"\n",
"# get speech timestamps from full audio file\n",
"speech_timestamps = get_speech_ts_adaptive(wav, model, run_function=validate_onnx) \n",
"pprint(speech_timestamps)"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:34:41.375446Z",
"start_time": "2021-04-15T13:34:41.368055Z"
},
"hidden": true
},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'save_audio' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m\u001b[0m",
"\u001b[0;31mNameError\u001b[0mTraceback (most recent call last)",
"\u001b[0;32m<ipython-input-5-713048adde74>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0;31m# merge all speech chunks to one audio\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0msave_audio\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'only_speech.wav'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcollect_chunks\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mspeech_timestamps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwav\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m16000\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 3\u001b[0m \u001b[0mAudio\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'only_speech.wav'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mNameError\u001b[0m: name 'save_audio' is not defined"
]
}
],
"source": [
"# merge all speech chunks to one audio\n",
"save_audio('only_speech.wav', collect_chunks(speech_timestamps, wav), 16000)\n",
"Audio('only_speech.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "Rio9W50gjsr1"
},
"source": [
"### Single Audio Stream"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:09.606031Z",
"start_time": "2020-12-15T13:09:09.504239Z"
},
"hidden": true,
"id": "IPkl8Yy1jsr1"
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"wav = f'{files_dir}/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"
},
"hidden": true,
"id": "NC6Jim0hjsr1"
},
"outputs": [],
"source": [
"for batch in single_audio_stream(model, wav, run_function=validate_onnx):\n",
" if batch:\n",
" pprint(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"wav = f'{files_dir}/en.wav'"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"for batch in single_audio_stream(model, wav, iterator_type='adaptive', run_function=validate_onnx):\n",
" if batch:\n",
" pprint(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "WNZ42u0ajsr1"
},
"source": [
"### Multiple Audio Streams"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:11.540423Z",
"start_time": "2020-12-15T13:09:11.455706Z"
},
"hidden": true,
"id": "XjhGQGppjsr1"
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"audios_for_stream = glob.glob(f'{files_dir}/*.wav')\n",
"pprint(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"
},
"hidden": true,
"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",
" pprint(batch)"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## Number detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "bL4kn4KJrlyL"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:25:19.107534Z",
"start_time": "2020-12-30T17:24:51.853293Z"
},
"cellView": "form",
"hidden": true,
"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 torch\n",
"import onnxruntime\n",
"from pprint import pprint\n",
"\n",
"from IPython.display import Audio\n",
"\n",
"_, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
" model='silero_number_detector',\n",
" force_reload=True)\n",
"\n",
"(get_number_ts,\n",
" save_audio,\n",
" read_audio,\n",
" collect_chunks,\n",
" drop_chunks) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'\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": {
"heading_collapsed": true,
"hidden": true,
"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"
},
"hidden": true,
"id": "krnGoA6Kjsr0"
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/number_detector.onnx')\n",
"wav = read_audio(f'{files_dir}/en_num.wav')\n",
"\n",
"# get number timestamps from full audio file\n",
"number_timestamps = get_number_ts(wav, model, run_function=validate_onnx)\n",
"pprint(number_timestamps)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"sample_rate = 16000\n",
"# convert ms in timestamps to samples\n",
"for timestamp in number_timestamps:\n",
" timestamp['start'] = int(timestamp['start'] * sample_rate / 1000)\n",
" timestamp['end'] = int(timestamp['end'] * sample_rate / 1000)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:08.862421Z",
"start_time": "2020-12-15T13:09:08.820014Z"
},
"hidden": true,
"id": "B176Lzfnjsr1"
},
"outputs": [],
"source": [
"# merge all number chunks to one audio\n",
"save_audio('only_numbers.wav',\n",
" collect_chunks(number_timestamps, wav), 16000) \n",
"Audio('only_numbers.wav')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"# drop all number chunks from audio\n",
"save_audio('no_numbers.wav',\n",
" drop_chunks(number_timestamps, wav), 16000) \n",
"Audio('no_numbers.wav')"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true
},
"source": [
"## Language detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "bL4kn4KJrlyL"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:25:19.107534Z",
"start_time": "2020-12-30T17:24:51.853293Z"
},
"cellView": "form",
"hidden": true,
"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 torch\n",
"import onnxruntime\n",
"from pprint import pprint\n",
"\n",
"from IPython.display import Audio\n",
"\n",
"_, utils = torch.hub.load(repo_or_dir='snakers4/silero-vad',\n",
" model='silero_lang_detector',\n",
" force_reload=True)\n",
"\n",
"(get_language,\n",
" read_audio) = utils\n",
"\n",
"files_dir = torch.hub.get_dir() + '/snakers4_silero-vad_master/files'\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": {
"hidden": true,
"id": "5JHErdB7jsr0"
},
"source": [
"### Full Audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/number_detector.onnx')\n",
"wav = read_audio(f'{files_dir}/en.wav')\n",
"\n",
"lang = get_language(wav, model, run_function=validate_onnx)\n",
"print(lang)"
]
}
],
"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": "3.8.8"
},
"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
}
Reference in New Issue View Git Blame Copy Permalink