yeyupiaoling/MASR
Pytorch实现的流式与非流式的自动语音识别框架，同时兼容在线和离线识别，目前支持Conformer、Squeezeformer、DeepSpeech2模型，支持多种数据增强方法。
Multiple decoding strategies (CTC greedy search, prefix beam search, attention rescoring) enable flexible accuracy-latency tradeoffs, while SentencePiece tokenization simplifies multilingual and mixed-language training. Deploys across server, Nvidia Jetson, and desktop platforms (Windows/Linux/macOS) with kaldi_native_fbank preprocessing for cross-platform performance optimization.
724 stars and 945 monthly downloads. Available on PyPI.
Stars
724
Forks
114
Language
Python
License
Apache-2.0
Category
Last pushed
Dec 17, 2025
Monthly downloads
945
Commits (30d)
0
Dependencies
18
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/yeyupiaoling/MASR"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
coqui-ai/TTS
🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production
netease-youdao/EmotiVoice
EmotiVoice 😊: a Multi-Voice and Prompt-Controlled TTS Engine
shivammehta25/Matcha-TTS
[ICASSP 2024] 🍵 Matcha-TTS: A fast TTS architecture with conditional flow matching
keithito/tacotron
A TensorFlow implementation of Google's Tacotron speech synthesis with pre-trained model (unofficial)
DigitalPhonetics/IMS-Toucan
Controllable and fast Text-to-Speech for over 7000 languages!