Files
hivecore_robot_voice/robot_speaker
NuoDaJia02 e5714e3a8b fix: Optimize voice interaction pipeline
1. register_speaker_node: Enable AEC to match main node for better SV accuracy.
2. tts/dashscope: Fix ffmpeg argument order (input option thread_queue_size).
3. asr/dashscope: Keep WebSocket connection alive to reduce latency.
4. speaker_verifier: Force single-thread inference to avoid CPU contention.
2026-01-19 16:17:27 +08:00
..
2026-01-19 09:58:40 +08:00
2026-01-19 14:32:57 +08:00
2026-01-12 17:40:08 +08:00