Hacktoberfest 2026: los issues que los mantenedores marcaron para octubre, abiertos y aptos para principiantes. Explorar issues de Hacktoberfest

Voice turn-taking that does not cut you off

Abierto
#168 0 comentarios 0 reacciones 0 asignados Ver en GitHub

Los mantenedores suelen responder en 1 día

Nadie ha tomado este issue todavía.

Evaluación

Dificultad
4/5
Tiempo estimado
3-5 días
Aptitud para principiantes
30/100
Tipo de issue
Nueva funcionalidad
Claridad
Bastante claro
Estado de actividad
Activo
Stack tecnológico
python

Línea de trabajo

Start with the end-of-turn logic in sentient/voice/vad.py and the vad_silence_ms setting in the voice section of sentient/config/schema.py, then trace how sentient/voice/session.py handles barge_in over /ws/voice in sentient/voice/socket.py. Done means pauses shorter than the configured silence keep the turn open, Sentient's own audio does not trigger barge-in, and a per-turn latency value is shown in desktop/src/features/voice/. Scope spans several subsystems and is a bundle of six ideas, so confirm with maintainers which item to take first.

Escrito por el modelo de indexación a partir del texto del issue.

Descripción

area: voice enhancement help wanted priority: P1 size: M

Voice failures are turn-taking failures. Alexa+ treats normal pauses as the end of a turn. Practitioners target under 800 ms end to end and find "not listening" worse than talking over the user. Gemini's redesigned mic "doesn't cut you off mid-thought", and Hermes added streaming STT.

Idea:

  • a tunable end-of-turn silence
  • echo cancellation so Sentient's own audio does not trigger barge-in
  • filtering of backchannels like "mhm"
  • streaming transcription
  • a latency readout per turn
  • typing and talking switch seamlessly in the same thread

Status in Sentient today: Partial. VAD with vad_silence_ms and interruptible /ws/voice with barge_in exist.

Where: sentient/voice/vad.py, sentient/voice/session.py, sentient/voice/stt.py, sentient/voice/socket.py, the voice section of sentient/config/schema.py, and desktop/src/features/voice/.

Done when:

  • pauses shorter than the setting do not end the turn
  • Sentient's own speech does not trigger barge-in
  • backchannels do not interrupt
  • latency is shown per turn

Sources:

Roadmap ID: VOICE-1 · Priority: P1 · Size: M

Lenguaje dominante
Python
Estrellas
710
Forks
125
Merge medio
8 h 8 min
PR fusionados (30 d)
180

Preparar el entorno

Primeros pasos

  1. Lee el issue completo y luego la guía de contribución del proyecto.
  2. Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
  3. Haz un fork del repositorio y trabaja en una rama.
  4. Abre un pull request que haga referencia al número del issue.

Más de existence-master/Sentient

Todos los issues de existence-master/Sentient

Issues similares

Más issues de Python

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.