livekit-plugins-alkove

Alkove for LiveKit Agents: Focus keeps the user's voice and removes the others and the noise, Cue finds where they speak. Both run on the agent's CPU, with no server and no key, with the chain and the speech rules of Alkove's web SDK.

Install

pip install livekit-plugins-alkove
python my_agent.py download-files     # Focus and Cue into the cache, checked by their sha256

Use

from livekit.agents import AgentSession, room_io
from livekit.plugins import alkove

session = AgentSession(vad=alkove.VAD.load(), ...)
await session.start(
    agent,
    room=ctx.room,
    room_options=room_io.RoomOptions(
        audio_input=room_io.AudioInputOptions(noise_cancellation=alkove.Focus()),
    ),
)

• alkove.Focus() is a rtc.FrameProcessor. It keeps the closest voice, or with voiceprint= the voice of that print; strength= and normalize= work as in the web SDK. It works at the session's sample rate (24 kHz by default) and gives each frame back at the size it came in, 20 ms behind: whole 10 ms frames, as LiveKit's gain control needs. The models are read from the cache download-files filled when it is created.
• alkove.VAD.load() gives one speech probability every 30 ms, with speech=alkove.SpeechParams(confidence=0.5, exit_confidence=0.1, start_secs=0.2, stop_secs=0.6) by default. It counts as the turn detector's min_silence_duration (stop_secs). Behind Focus, it hears only the user.

The models are loaded once per process and shared by every session. Pass models=alkove.Models.load(["focus", "cue"], dir=...) to read them from your own folder.

Tested with livekit-agents 1.8.8.

License

Apache-2.0. See NOTICE.