Open Source
Explore the latest AI open-source projects from GitHub and HuggingFace.
Explore the latest AI open-source projects from GitHub and HuggingFace.
FunASR is a fundamental end-to-end speech recognition toolkit developed by Alibaba DAMO Academy (ModelScope), bridging academic research and industrial applications. It provides 20+ industrial-grade pretrained models including Paraformer, SenseVoice, and Fun-ASR-Nano, supporting both streaming and non-streaming ASR, voice activity detection (VAD), punctuation restoration, speaker diarization, emotion recognition, and keyword spotting. Fun-ASR-Nano supports 31 languages with low-latency real-time transcription trained on tens of millions of hours of real speech data.
ggml-org
Pure C/C++ port of OpenAI Whisper for edge deployment
SYSTRAN
High-performance Whisper reimplementation using CTranslate2, delivering 4x faster speech recognition with INT8 quantization support.
m-bain
Fast Whisper-based ASR with word-level timestamps and multi-speaker diarization, 70x real-time speed
KoljaB
A robust, low-latency Python library for real-time speech-to-text with integrated voice activity detection, wake word activation, and Faster Whisper transcription.