A Repository for Single- and Multi-modal Speaker Verification, Speaker Recognition and Speaker Diarization

vits2 backbone with multilingual-bert

RWTH Aachen University, Germany(Hermann Ney)

A PyTorch-based Speech Toolkit

We Speech Transcript based on LLM, in 300 lines of code.

(NOT Official) Implementation Auto-regressive Integrate-and-Fire (AIF)

Implementation of BEST-RQ - a model for self-supervised learning of speech signals using a random projection quantizer, in Pytorch.

Fine-tune and evaluate Whisper models for Automatic Speech Recognition (ASR) on custom datasets or datasets from huggingface.

Standalone implementation of the CUDA-accelerated WFST Decoder available in Riva

S2ST 伪标签

S2ST Data

Computes the MWER (minimum WER) Loss with CTC beam search. Knowledge distillation for CTC loss.

Neural end-to-end Speech Translation Toolkit

gfcc features

End-to-end MOdeling of ASR (Automatic Speech Recognition)

Keyword spotting, Speech wake_up, by pytorch, DNN, CNN, TDNN, DFSMN, LSTM

对常用的6700个汉字进行音、形比较，输出音近字、形近字的列表。 # 相近字

Pre-processing KsponSpeech corpus (Korean Speech dataset) provided by AI Hub.

Unsupervised text tokenizer for Neural Network-based text generation.

TensorFlow code and pre-trained models for BERT

This repository contains the Kaldi LF-MMI implementation of the paper "Bayesian Learning of LF-MMI Trained Time Delay Neural Networks for Speech Recognition"

从Kaldi中裁剪的轻量级语音识别解码推理框架，目前实现了MFCC+GMM+Viterbi，不依赖OpenFST、OpenBLAS等库