Feature Request
Add FunASR as a voice-to-note transcription engine for NoteGen.
Why
Voice notes are a natural input method for note-taking. FunASR provides:
- SenseVoice (234M): 50+ languages, 25x realtime on CPU — fast enough for real-time dictation
- Built-in punctuation restoration: Clean, readable transcription output ready for Markdown
- Emotion detection: Tag notes with detected mood
- Fully offline: No cloud API, privacy-friendly
- OpenAI-compatible API:
funasr-server at POST /v1/audio/transcriptions
References
Feature Request
Add FunASR as a voice-to-note transcription engine for NoteGen.
Why
Voice notes are a natural input method for note-taking. FunASR provides:
funasr-serveratPOST /v1/audio/transcriptionsReferences