zhu-gu-an Goto Github PK

3d-speaker

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

asr-decode

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

asr-decoders

Deployable end-to-end asr decoders

asr_vad

audio-distance-estimation

audio-preprocess

Preprocess Audio for training

audio-preprocessing-scripts

语音数据集自动化制作脚本

ava-speech-downloader

AVA-Speech dataset downloader

awesome-chinese-llm

整理开源的中文大语言模型，以规模较小、可私有化部署、训练成本较低的模型为主，包括底座模型，垂直领域微调及应用，数据集与教程等。

awesome-llms-meet-multimodal-generation

🔥🔥🔥 A curated list of papers on LLMs-based multimodal generation (image, video, 3D and audio).

batch-whisper

Batch Support for OpenAI Whisper

bert-ner-pytorch

Chinese NER(Named Entity Recognition) using BERT(Softmax, CRF, Span)

brouhaha-vad

Predicts the level of noise and reverberation on your audiofiles

capswriter-offline

CapsWriter 简陋但好用的离线版，一个 PC 端的语音输入工具

chattts-ui

一个简单的本地网页界面，直接使用ChatTTS将文字合成为语音，同时支持对外提供API接口。

chinese_speech_pretrain

chinese speech pretrained models

chinesenlpcorpus

中文自然语言处理数据集，平时做做实验的材料。欢迎补充提交合并。

coat

(ICCV 2021 Oral) CoaT: Co-Scale Conv-Attentional Image Transformers

computepitchfeature

A modification from Kaldi's ComputePitchFeats.

cpp_new_features

2021年最新整理， C++ 学习资料，含C++ 11 / 14 / 17 / 20 / 23 新特性、入门教程、推荐书籍、优质文章、学习笔记、教学视频等

ctc_decoder

A ctc decoder for both online and offline asr model

cuda-learn-note

🎉CUDA 笔记 / 大模型手撕CUDA / C++笔记，更新随缘: flash_attn、sgemm、sgemv、warp reduce、block reduce、dot product、elementwise、softmax、layernorm、rmsnorm、hist etc.

damo-fsmn-vad-infer-httpserver

达摩fsmn vad c++推理服务

deepfilternet

Noise supression using deep filtering

deepseek-llm

DeepSeek LLM: Let there be answers

e2e_asr_confidence_estimation

Implementation of the paper "Confidence estimation for attention based sequence to sequence models for speech recognition"

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

fastasr

这是一个用C++实现ASR推理的项目，它依赖很少，安装也很简单，推理速度很快，在树莓派4B等ARM平台也可以流畅的运行。 支持的模型是由Google的Transformer模型中优化而来，数据集是开源wenetspeech(10000+小时)或阿里私有数据集(60000+小时)， 所以识别效果也很好，可以媲美许多商用的ASR软件。

funasr

A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models.

funnlp

NLP文本数据