Alibaba Tongyi Lab releases open source voice large model project FunAudioLLM

Alibaba Tongyi Lab recently released the open source voice model project FunAudioLLM, which includes two models: SenseVoice and CosyVoice. SenseVoice focuses on high-precision multi-language speech recognition, emotion recognition, and audio event detection, supporting more than 50 languages, and its effect is better than the Whisper model. CosyVoice focuses on natural speech generation, supporting multiple languages, timbre, and emotion control.