Vad-sli-asr - A Python scripts for a speech processing pipeline with Voice Activity Detection (VAD)

Last update: Dec 09, 2022

Related tags

Overview

VAD-SLI-ASR

Python scripts for a speech processing pipeline with Voice Activity Detection (VAD), Spoken Language Identification (SLI), and Automatic Speech Recognition (ASR). Our use case involves using VAD to detect time regions in a language documentation recording where someone is speaking, then using SLI to classify each region as either English (eng) or Muruwari (zmu), and then using an English ASR model to transcribe regions detected as English. This pipeline outputs an ELAN .eaf file with the following tier structure (_vad, _sli, and _asr):

Set up

pip install -r requirements.txt

Data

├── data
│   ├── sli-train      <- Training data for SLI (one folder per language)
│   │   ├── eng/       <- .wav files (English utterances)
│   │   ├── zmu/       <- .wav files (Muruwari utterances)
│   ├── asr-train      <- Intermediate data that has been transformed.
│   │   ├── eng.tsv    <- transcriptions
│   │   ├── eng/       <- .wav files (English utterances)

Usage

VAD

# VAD
python scripts/run_vad-by-silero.py myrecording.wav

SLI

# To train a classifier using your own clips and then save it:
python scripts/train_sli-by-sblr.py data/sli-train models/zmu-eng_sli_k10.pkl

# Use trained model to classify VAD-detected regions as eng or zmu
python scripts/run_sli-by-sblr.py models/zmu-eng_sli_k10.pkl myrecording.wav

ASR

# To fine-tune a wav2vec 2.0 model and save the checkpoint:
python scripts/train_asr-by-w2v2.py data/asr-train data/checkpoints/no-lm_b10

# Transcribe using trained model 
python scripts/run_asr-by-w2v2.py data/checkpoints/no-lm_b10 myrecording.wav

Paddlespeech Streaming ASR GUI

Paddlespeech-Streaming-ASR-GUI Introduction A paddlespeech Streaming ASR GUI. Us

3 Jan 5, 2022

Code for Findings of ACL 2022 Paper "Sentiment Word Aware Multimodal Refinement for Multimodal Sentiment Analysis with ASR Errors"

SWRM Code for Findings of ACL 2022 Paper "Sentiment Word Aware Multimodal Refinement for Multimodal Sentiment Analysis with ASR Errors" Clone Clone th

14 Jan 3, 2023

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

Rasa Open Source Rasa is an open source machine learning framework to automate text-and voice-based conversations. With Rasa, you can build contextual

15.3k Dec 30, 2022

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

Rasa Open Source Rasa is an open source machine learning framework to automate text-and voice-based conversations. With Rasa, you can build contextual

15.3k Jan 3, 2023

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

Rasa Open Source Rasa is an open source machine learning framework to automate text-and voice-based conversations. With Rasa, you can build contextual

10.8k Feb 18, 2021

Releases(1.1.0)

1.1.0(Apr 23, 2022)
Switched to using pre-existing vocabulary from pre-trained model (see Appendix A in paper).

Source code(tar.gz)
Source code(zip)
1.0.0(Apr 18, 2022)

Source code(tar.gz)
Source code(zip)
0.9.0(Apr 14, 2022)

Pre-release to check Zenodo sync
Source code(tar.gz)
Source code(zip)

Vad-sli-asr - A Python scripts for a speech processing pipeline with Voice Activity Detection (VAD)

Related tags

Overview

VAD-SLI-ASR

Set up

Data

Usage

VAD

SLI

ASR

You might also like...

Paddlespeech Streaming ASR GUI

Code for Findings of ACL 2022 Paper "Sentiment Word Aware Multimodal Refinement for Multimodal Sentiment Analysis with ASR Errors"

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

This project converts your human voice input to its text transcript and to an automated voice too.

Chinese real time voice cloning (VC) and Chinese text to speech (TTS).

Clone a voice in 5 seconds to generate arbitrary speech in real-time

Every Google, Azure & IBM text to speech voice for free

Releases(1.1.0)

1.1.0(Apr 23, 2022)

1.0.0(Apr 18, 2022)

0.9.0(Apr 14, 2022)

Owner

Dynamics of Language

edge-SR: Super-Resolution For The Masses

Snips Python library to extract meaning from text

MicBot - MicBot uses Google Translate to speak everyone's chat messages

👑 spaCy building blocks and visualizers for Streamlit apps

PyTorch implementation and pretrained models for XCiT models. See XCiT: Cross-Covariance Image Transformer

An official implementation for "CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retrieval"

FewCLUE: 为中文NLP定制的小样本学习测评基准

Reproduction process of BERT on SST2 dataset

Yet Another Sequence Encoder - Encode sequences to vector of vector in python !

Neural text generators like the GPT models promise a general-purpose means of manipulating texts.

The proliferation of disinformation across social media has led the application of deep learning techniques to detect fake news.

Modular and extensible speech recognition library leveraging pytorch-lightning and hydra.

sangha, pronounced "suhng-guh", is a social networking, booking platform where students and teachers can share their practice.

In this repository we have tested 3 VQA models on the ImageCLEF-2019 dataset.

Include MelGAN, HifiGAN and Multiband-HifiGAN, maybe NHV in the future.

Speech Recognition Database Management with python

SIGIR'22 paper: Axiomatically Regularized Pre-training for Ad hoc Search

A notebook that shows how to import the IITB English-Hindi Parallel Corpus from the HuggingFace datasets repository

A Streamlit web app that generates Rick and Morty stories using GPT2.

This project is part of Eleuther AI's quest to create a massive repository of high quality text data for training language models.