Article may be outdated

This article is 5 days old. Some details may have changed since publication.

Hacker News·4 min read·hard

Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech

K
koljab
Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech
✦AI Summary

RealtimeVAD is a new tool designed to reduce latency in voice assistants by detecting the end of speech faster than standard models like Silero. It uses a dual-check system to identify silence more quickly, improving the responsiveness of voice-based AI interactions.

Why it matters

Reducing latency is critical for making AI voice assistants feel more natural and conversational.

✦Dive DeeperCreate a free account to unlock

Know that the speaker stopped, about 100 ms before Silero does.

RealtimeSTT · RealtimeTTS · Live demo · Spec · LiveKit plugin

A voice assistant can't do anything until it knows the user has stopped talking. Speech recognition, the LLM call and the voice output all wait for that one decision, and a normal voice activity detector takes a few hundred milliseconds to make it.

RealtimeVAD makes it sooner. It keeps Silero VAD to decide what is speech, and adds a second, faster check: every 10 ms it compares the audio level with the room's noise and with the speaker's own level. The moment the audio goes quiet, you get an event, long before Silero's probability has fallen.

demo.mp4 The video uses a synthetic voice and shows a median lead of 92 ms. The 100 ms above is the median on real voice-assistant turns.

Continue reading on Headlinne

Create a free account to read the full article.

Read full article →
technologyai
✦

Get smarter about the news

Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.

Create free account

Already have an account? Sign in