deepgeek/silero-vad
2
0
Fork 0
mirror of https://github.com/snakers4/silero-vad.git synced 2026-02-05 09:59:20 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
ec16a93fc451a14ec3c584769d9225e458a57076
silero-vad/silero-vad.ipynb
adamnsandle ec16a93fc4 fx notebook
2021-04-15 14:08:49 +00:00

1398 lines
55 KiB
Plaintext
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,
"id": "FpMplOCA2Fwp"
},
"source": [
"## VAD"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "62A6F_072Fwq"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-30T17:35:43.397137Z",
"start_time": "2020-12-30T17:33:10.962078Z"
},
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "5w5AkskZ2Fwr",
"outputId": "545c0988-965d-4462-eb06-d4c5a48d8969"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[?25l\r\u001b[K |▏ | 10kB 16.5MB/s eta 0:00:01\r\u001b[K |▍ | 20kB 20.7MB/s eta 0:00:01\r\u001b[K |▌ | 30kB 23.6MB/s eta 0:00:01\r\u001b[K |▊ | 40kB 24.5MB/s eta 0:00:01\r\u001b[K |▉ | 51kB 25.9MB/s eta 0:00:01\r\u001b[K |█ | 61kB 23.6MB/s eta 0:00:01\r\u001b[K |█▏ | 71kB 19.5MB/s eta 0:00:01\r\u001b[K |█▍ | 81kB 20.3MB/s eta 0:00:01\r\u001b[K |█▌ | 92kB 18.4MB/s eta 0:00:01\r\u001b[K |█▊ | 102kB 17.6MB/s eta 0:00:01\r\u001b[K |█▉ | 112kB 17.6MB/s eta 0:00:01\r\u001b[K |██ | 122kB 17.6MB/s eta 0:00:01\r\u001b[K |██▏ | 133kB 17.6MB/s eta 0:00:01\r\u001b[K |██▍ | 143kB 17.6MB/s eta 0:00:01\r\u001b[K |██▌ | 153kB 17.6MB/s eta 0:00:01\r\u001b[K |██▊ | 163kB 17.6MB/s eta 0:00:01\r\u001b[K |██▉ | 174kB 17.6MB/s eta 0:00:01\r\u001b[K |███ | 184kB 17.6MB/s eta 0:00:01\r\u001b[K |███▏ | 194kB 17.6MB/s eta 0:00:01\r\u001b[K |███▍ | 204kB 17.6MB/s eta 0:00:01\r\u001b[K |███▌ | 215kB 17.6MB/s eta 0:00:01\r\u001b[K |███▊ | 225kB 17.6MB/s eta 0:00:01\r\u001b[K |███▉ | 235kB 17.6MB/s eta 0:00:01\r\u001b[K |████ | 245kB 17.6MB/s eta 0:00:01\r\u001b[K |████▏ | 256kB 17.6MB/s eta 0:00:01\r\u001b[K |████▍ | 266kB 17.6MB/s eta 0:00:01\r\u001b[K |████▌ | 276kB 17.6MB/s eta 0:00:01\r\u001b[K |████▊ | 286kB 17.6MB/s eta 0:00:01\r\u001b[K |█████ | 296kB 17.6MB/s eta 0:00:01\r\u001b[K |█████ | 307kB 17.6MB/s eta 0:00:01\r\u001b[K |█████▎ | 317kB 17.6MB/s eta 0:00:01\r\u001b[K |█████▍ | 327kB 17.6MB/s eta 0:00:01\r\u001b[K |█████▋ | 337kB 17.6MB/s eta 0:00:01\r\u001b[K |█████▊ | 348kB 17.6MB/s eta 0:00:01\r\u001b[K |██████ | 358kB 17.6MB/s eta 0:00:01\r\u001b[K |██████ | 368kB 17.6MB/s eta 0:00:01\r\u001b[K |██████▎ | 378kB 17.6MB/s eta 0:00:01\r\u001b[K |██████▍ | 389kB 17.6MB/s eta 0:00:01\r\u001b[K |██████▋ | 399kB 17.6MB/s eta 0:00:01\r\u001b[K |██████▊ | 409kB 17.6MB/s eta 0:00:01\r\u001b[K |███████ | 419kB 17.6MB/s eta 0:00:01\r\u001b[K |███████ | 430kB 17.6MB/s eta 0:00:01\r\u001b[K |███████▎ | 440kB 17.6MB/s eta 0:00:01\r\u001b[K |███████▍ | 450kB 17.6MB/s eta 0:00:01\r\u001b[K |███████▋ | 460kB 17.6MB/s eta 0:00:01\r\u001b[K |███████▊ | 471kB 17.6MB/s eta 0:00:01\r\u001b[K |████████ | 481kB 17.6MB/s eta 0:00:01\r\u001b[K |████████ | 491kB 17.6MB/s eta 0:00:01\r\u001b[K |████████▎ | 501kB 17.6MB/s eta 0:00:01\r\u001b[K |████████▍ | 512kB 17.6MB/s eta 0:00:01\r\u001b[K |████████▋ | 522kB 17.6MB/s eta 0:00:01\r\u001b[K |████████▊ | 532kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████ | 542kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████ | 552kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████▎ | 563kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████▌ | 573kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████▋ | 583kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████▉ | 593kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████ | 604kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████▏ | 614kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████▎ | 624kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████▌ | 634kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████▋ | 645kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████▉ | 655kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████ | 665kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████▏ | 675kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████▎ | 686kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████▌ | 696kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████▋ | 706kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████▉ | 716kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████ | 727kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████▏ | 737kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████▎ | 747kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████▌ | 757kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████▋ | 768kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████▉ | 778kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████ | 788kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████▏ | 798kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████▎ | 808kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████▌ | 819kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████▋ | 829kB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████▉ | 839kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████ | 849kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████▏ | 860kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████▍ | 870kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████▌ | 880kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████▊ | 890kB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████▉ | 901kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████ | 911kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████▏ | 921kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████▍ | 931kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████▌ | 942kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████▊ | 952kB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████▉ | 962kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████ | 972kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████▏ | 983kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████▍ | 993kB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████▌ | 1.0MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████▊ | 1.0MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████▉ | 1.0MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████ | 1.0MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████▏ | 1.0MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████▍ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████▌ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████▊ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████▉ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████▏ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████▍ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████▌ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████▊ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████ | 1.1MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████▎ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████▍ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████▋ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████▊ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████▎ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████▍ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████▋ | 1.2MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████▊ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████▎ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████▍ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████▋ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████▊ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████ | 1.3MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████▎ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████▍ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████▋ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████▊ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████▎ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████▌ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████▋ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████▉ | 1.4MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████▏ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████▎ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████▌ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████▋ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████▉ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████▏ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████▎ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████▌ | 1.5MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████▋ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████▉ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████▏ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████▎ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████▌ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████▋ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████▉ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████▏ | 1.6MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████▎ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████▌ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████▋ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████▉ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████▏ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████▍ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████▌ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████▊ | 1.7MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████▉ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████▏ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████▍ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████▌ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████▊ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |█████████████████████████████▉ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▏ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▍ | 1.8MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▌ | 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▊ | 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▉ | 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████ | 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▏| 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▍| 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▌| 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▊| 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▉| 1.9MB 17.6MB/s eta 0:00:01\r\u001b[K |████████████████████████████████| 1.9MB 17.6MB/s \n",
"\u001b[?25h"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Downloading: \"https://github.com/snakers4/silero-vad/archive/master.zip\" to /root/.cache/torch/hub/master.zip\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",
"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,
"id": "dY2Us3_Q2Fws"
},
"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": "2020-12-30T17:35:44.362860Z",
"start_time": "2020-12-30T17:35:43.398441Z"
},
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "aI_eydBPjsrx",
"outputId": "17d317e6-ec8c-46a2-c5ec-682c1391e58d"
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py:889: UserWarning: stft will soon require the return_complex parameter be given for real inputs, and will further require that return_complex=True in a future PyTorch release. (Triggered internally at /pytorch/aten/src/ATen/native/SpectralOps.cpp:639.)\n",
" result = self.forward(*input, **kwargs)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'end': 35000, 'start': 0},\n",
" {'end': 112000, 'start': 35000},\n",
" {'end': 124000, 'start': 112000},\n",
" {'end': 320000, 'start': 143000},\n",
" {'end': 628000, 'start': 319000},\n",
" {'end': 752000, 'start': 632000},\n",
" {'end': 801000, 'start': 775000},\n",
" {'end': 960000, 'start': 811000}]\n"
]
}
],
"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,
"id": "n8plzbJU2Fws"
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "SQOtu2Vl2Fwt",
"outputId": "3a560cf3-a882-4db7-ad7e-0ab9bf1a9698"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'end': 35000, 'start': 0},\n",
" {'end': 112000, 'start': 35500},\n",
" {'end': 246000, 'start': 142500},\n",
" {'end': 288500, 'start': 251500},\n",
" {'end': 315500, 'start': 289500},\n",
" {'end': 603500, 'start': 318000},\n",
" {'end': 623000, 'start': 606500},\n",
" {'end': 713000, 'start': 631000},\n",
" {'end': 728500, 'start': 712000},\n",
" {'end': 748500, 'start': 726500},\n",
" {'end': 798500, 'start': 775000},\n",
" {'end': 899500, 'start': 811000},\n",
" {'end': 914000, 'start': 897000},\n",
" {'end': 962000, 'start': 913000}]\n"
]
}
],
"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,
"id": "Lr6zCGXh2Fwt"
},
"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,
"id": "xCM-HrUR2Fwu"
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:59.199321Z",
"start_time": "2020-12-15T13:09:59.196823Z"
},
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "q-lql_2Wjsry",
"outputId": "ada632d4-eaba-475e-b00c-fa8238411792"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{4000: 'start'}]\n",
"[{39000: 'end'}]\n",
"[{43000: 'start'}]\n",
"[{115500: 'end'}]\n",
"[{121500: 'start'}]\n",
"[{127500: 'end'}]\n",
"[{150500: 'start'}]\n",
"[{291000: 'end'}]\n",
"[{295000: 'start'}]\n",
"[{322000: 'end'}]\n",
"[{326500: 'start'}]\n",
"[{631500: 'end'}]\n",
"[{640500: 'start'}]\n",
"[{755000: 'end'}]\n",
"[{782500: 'start'}]\n",
"[{804500: 'end'}]\n",
"[{818500: 'start'}]\n"
]
}
],
"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,
"id": "t8TXtnvk2Fwv"
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "BX3UgwwB2Fwv",
"outputId": "8d704639-6f3e-4520-d6ac-7ac988265286"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{2000: 'start'}]\n",
"[{40000: 'end'}]\n",
"[{44000: 'start'}]\n",
"[{115500: 'end'}]\n",
"[{151000: 'start'}]\n",
"[{251000: 'end'}]\n",
"[{260000: 'start'}]\n",
"[{291500: 'end'}]\n",
"[{298000: 'start'}]\n",
"[{320500: 'end'}]\n",
"[{326500: 'start'}]\n",
"[{612500: 'end'}]\n",
"[{615000: 'start'}]\n",
"[{628000: 'end'}]\n",
"[{639500: 'start'}]\n",
"[{718500: 'end'}]\n",
"[{720500: 'start'}]\n",
"[{755500: 'end'}]\n",
"[{783500: 'start'}]\n",
"[{805000: 'end'}]\n",
"[{819500: 'start'}]\n",
"[{902000: 'end'}]\n",
"[{905500: 'start'}]\n",
"[{921000: 'start'}]\n"
]
}
],
"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,
"id": "36jY0niD2Fww"
},
"source": [
"## Number detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "scd1DlS42Fwx"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true,
"id": "Kq5gQuYq2Fwx"
},
"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,
"id": "qhPa30ij2Fwy"
},
"source": [
"### Full audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true,
"id": "EXpau6xq2Fwy"
},
"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,
"id": "u-KfXRhZ2Fwy"
},
"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,
"id": "iwYEC4aZ2Fwy"
},
"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,
"id": "fHaYejX12Fwy"
},
"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,
"id": "PnKtJKbq2Fwz"
},
"source": [
"## Language detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "F5cAmMbP2Fwz"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true,
"id": "Zu9D0t6n2Fwz"
},
"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,
"id": "iC696eMX2Fwz"
},
"source": [
"### Full audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true,
"id": "c8UYnYBF2Fw0"
},
"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,
"id": "hEhnfORV2Fw0"
},
"source": [
"## VAD"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "bL4kn4KJrlyL"
},
"source": [
"### Install Dependencies"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:30:22.938755Z",
"start_time": "2021-04-15T13:30:20.970574Z"
},
"cellView": "form",
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "Q4QIfSpprnkI",
"outputId": "119e85c9-bb9a-43bb-ae23-7d197b470096"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[K |████████████████████████████████| 4.1MB 19.5MB/s \n",
"\u001b[?25h"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Downloading: \"https://github.com/snakers4/silero-vad/archive/master.zip\" to /root/.cache/torch/hub/master.zip\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 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,
"id": "TNEtK5zi2Fw2"
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:30:14.475412Z",
"start_time": "2021-04-15T13:30:14.427933Z"
},
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "krnGoA6Kjsr0",
"outputId": "edab010a-e066-42a0-9b4c-2ab2579b6b47"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'end': 33000, 'start': 0},\n",
" {'end': 112000, 'start': 35000},\n",
" {'end': 287000, 'start': 143000},\n",
" {'end': 317000, 'start': 287000},\n",
" {'end': 623000, 'start': 319000},\n",
" {'end': 752000, 'start': 632000},\n",
" {'end': 801000, 'start': 775000},\n",
" {'end': 960000, 'start': 811000}]\n"
]
}
],
"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,
"id": "21RE8KEC2Fw2"
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "uIVs56rb2Fw2",
"outputId": "50ce9117-17d8-4bef-eb53-7204c56c4b7b"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{'end': 35000, 'start': 0},\n",
" {'end': 112500, 'start': 34500},\n",
" {'end': 245000, 'start': 140000},\n",
" {'end': 286500, 'start': 251500},\n",
" {'end': 315000, 'start': 285000},\n",
" {'end': 527500, 'start': 316500},\n",
" {'end': 603500, 'start': 524500},\n",
" {'end': 623500, 'start': 606500},\n",
" {'end': 713000, 'start': 629500},\n",
" {'end': 738500, 'start': 711500},\n",
" {'end': 751000, 'start': 735000},\n",
" {'end': 797500, 'start': 772500},\n",
" {'end': 883000, 'start': 809000},\n",
" {'end': 914500, 'start': 897000},\n",
" {'end': 962000, 'start': 911500}]\n"
]
}
],
"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": null,
"metadata": {
"ExecuteTime": {
"end_time": "2021-04-15T13:34:41.375446Z",
"start_time": "2021-04-15T13:34:41.368055Z"
},
"hidden": true,
"id": "cox6oumC2Fw3",
"outputId": "1c18d7b1-ae80-42cb-c2e1-b6494104e5f7"
},
"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,
"id": "i8EZwtaA2Fw3"
},
"source": [
"**Classic way of getting speech chunks, you may need to select the tresholds yourself**"
]
},
{
"cell_type": "code",
"execution_count": 9,
"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": 10,
"metadata": {
"ExecuteTime": {
"end_time": "2020-12-15T13:09:11.453171Z",
"start_time": "2020-12-15T13:09:09.633435Z"
},
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "NC6Jim0hjsr1",
"outputId": "4c48843d-8510-4d26-c546-220e22a85361"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{4000: 'start'}]\n",
"[{37000: 'end'}]\n",
"[{43000: 'start'}]\n",
"[{115500: 'end'}]\n",
"[{150500: 'start'}]\n",
"[{291000: 'end'}]\n",
"[{294500: 'start'}]\n",
"[{321000: 'end'}]\n",
"[{326500: 'start'}]\n",
"[{627000: 'end'}]\n",
"[{639000: 'start'}]\n",
"[{718000: 'end'}]\n",
"[{721000: 'start'}]\n",
"[{755500: 'end'}]\n",
"[{783000: 'start'}]\n",
"[{804000: 'end'}]\n",
"[{818500: 'start'}]\n"
]
}
],
"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,
"id": "0pSKslpz2Fw3"
},
"source": [
"**Experimental Adaptive method, algorythm selects tresholds itself (see readme for more information)**"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"hidden": true,
"id": "RZwc-Khk2Fw4"
},
"outputs": [],
"source": [
"model = init_onnx_model(f'{files_dir}/model.onnx')\n",
"wav = f'{files_dir}/en.wav'"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"hidden": true,
"id": "Z4lzFPs02Fw4",
"outputId": "8d2f9cb4-dbc7-4c7c-dde1-ff3b3297aa07"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{0: 'start'}]\n",
"[{38000: 'end'}]\n",
"[{43000: 'start'}]\n",
"[{115000: 'end'}]\n",
"[{148500: 'start'}]\n",
"[{250500: 'end'}]\n",
"[{260000: 'start'}]\n",
"[{292000: 'end'}]\n",
"[{293500: 'start'}]\n",
"[{320000: 'end'}]\n",
"[{325000: 'start'}]\n",
"[{548000: 'end'}]\n",
"[{547500: 'start'}]\n",
"[{613000: 'end'}]\n",
"[{615000: 'start'}]\n",
"[{626500: 'end'}]\n",
"[{638000: 'start'}]\n",
"[{697500: 'start'}]\n",
"[{718000: 'end'}]\n",
"[{720000: 'start'}]\n",
"[{756000: 'end'}]\n",
"[{781000: 'start'}]\n",
"[{804500: 'end'}]\n",
"[{817500: 'start'}]\n",
"[{872000: 'end'}]\n",
"[{871000: 'start'}]\n",
"[{902000: 'end'}]\n",
"[{905500: 'start'}]\n",
"[{920500: 'end'}]\n",
"[{920000: 'start'}]\n"
]
}
],
"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,
"id": "7QMvUvpg2Fw4"
},
"source": [
"## Number detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "tBPDkpHr2Fw4"
},
"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": "PdjGd56R2Fw5"
},
"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": "I9QWSFZh2Fw5"
},
"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": "_r6QZiwu2Fw5"
},
"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,
"id": "FN4aDwLV2Fw5"
},
"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": "JnvS6WTK2Fw5"
},
"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,
"id": "yUxOcOFG2Fw6"
},
"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,
"id": "SR8Bgcd52Fw6"
},
"source": [
"## Language detector"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
"hidden": true,
"id": "PBnXPtKo2Fw6"
},
"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": "iNkDWJ3H2Fw6"
},
"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": "G8N8oP4q2Fw6"
},
"source": [
"### Full Audio"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"hidden": true,
"id": "WHXnh9IV2Fw6"
},
"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": 0
}
Reference in New Issue View Git Blame Copy Permalink