ASR Latency
--
ms
TTS First Chunk
--
ms
TTS Total
--
ms
End-to-End
--
ms
AEC
NLMS Adaptive
Echo Cancel
VAD
Silero VAD
Voice Activity
Diarization
Pyannote
Speaker ID
Gate
Primary Speaker
Audio Filter
ASR
Whisper
Speech-to-Text
TTS
OmniVoice
Text-to-Speech
Event Log
Click "Connect" to start.
Transcript
Audio Tools
About
- Modular pipeline: AEC (NLMS) → VAD (Silero) → Diarization (Pyannote) → Gate (Primary Speaker) → ASR (Whisper) → TTS (OmniVoice)
- All blocks are hot-swappable at runtime
- Green dot = speech detected, gray = silence
- Click Skip to cancel current playback