Trendoday

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

discuss

  • Points26
  • Comments5
–Rank now
#16Peak rank
2Days on list
55Score · 7d
Nov 12, 2026First seen
Rank · 7 days

On Hacker News

Hey guys. I recently watched the latency reports for my local voice assistant, which show a waveform plus silero probability for voice activity. I noticed silero lags visibly on the speech off signal. There had to be a method to get this signal closer to what I felt is optimal. My idea was to use a 10 ms energy detector with a threshold that adapts to environment or room noise. Which worked amazingly well. It created more false positives but in my speculative generation pipeline that would not…

Points
26
Comments
5
Submitted by
koljab
Posted
Nov 12, 2026
Site
github.com

Story details from Hacker News ↗ · linked article on github.com.

Trending history

Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech was first seen on Hacker News · Front page on Nov 12, 2026. Since then it has appeared in 73 checks on 2 different days, across 2 lists (Hacker News · Front page, Hacker News · Show HN). Its best position was #16 of 30, reached on Nov 13, 05:13 UTC. It dropped off the list after Nov 13, 21:22 UTC.

Rank history

Position on Hacker News · Front page at every check (lower is better).

#1#11#20#30 03:5805:3607:1308:5110:28 Nov 13, 03:58 UTC · #23Nov 13, 04:13 UTC · #25Nov 13, 04:28 UTC · #18Nov 13, 04:43 UTC · #23Nov 13, 04:58 UTC · #25Nov 13, 05:13 UTC · #16Nov 13, 05:28 UTC · #18Nov 13, 05:43 UTC · #20Nov 13, 05:58 UTC · #22Nov 13, 06:13 UTC · #20Nov 13, 06:28 UTC · #19Nov 13, 06:43 UTC · #21Nov 13, 06:58 UTC · #21Nov 13, 07:13 UTC · #24Nov 13, 07:28 UTC · #26Nov 13, 07:43 UTC · #27Nov 13, 07:58 UTC · #21Nov 13, 08:13 UTC · #22Nov 13, 08:28 UTC · #22Nov 13, 08:43 UTC · #23Nov 13, 08:58 UTC · #23Nov 13, 09:13 UTC · #24Nov 13, 09:28 UTC · #25Nov 13, 09:43 UTC · #25Nov 13, 09:58 UTC · #25Nov 13, 10:13 UTC · #27Nov 13, 10:28 UTC · #30

Last 14 days

Best rank per day (UTC).

Nov 3today

Metric history

Platform metric as reported by Hacker News.

081725 03:5805:3607:1308:5110:28 Nov 13, 03:58 UTC · 5Nov 13, 04:13 UTC · 5Nov 13, 04:28 UTC · 6Nov 13, 04:43 UTC · 7Nov 13, 04:58 UTC · 8Nov 13, 05:13 UTC · 10Nov 13, 05:28 UTC · 10Nov 13, 05:43 UTC · 12Nov 13, 05:58 UTC · 12Nov 13, 06:13 UTC · 13Nov 13, 06:28 UTC · 14Nov 13, 06:43 UTC · 15Nov 13, 06:58 UTC · 17Nov 13, 07:13 UTC · 17Nov 13, 07:28 UTC · 17Nov 13, 07:43 UTC · 18Nov 13, 07:58 UTC · 20Nov 13, 08:13 UTC · 21Nov 13, 08:28 UTC · 22Nov 13, 08:43 UTC · 23Nov 13, 08:58 UTC · 23Nov 13, 09:13 UTC · 23Nov 13, 09:28 UTC · 23Nov 13, 09:43 UTC · 23Nov 13, 09:58 UTC · 23Nov 13, 10:13 UTC · 24Nov 13, 10:28 UTC · 25

Where it trended

ListBestDaysChecksLast seen
Hacker News · Front page#16127
Hacker News · Show HN#1246

From the discussion

A few top-level comments from Hacker News, shortened. Read the full thread on Hacker News.

You can also avoid the problem all together. I made a "speculative regeneration" at least thats what i call it pipeline i been using for many months now and it works great. basically as long as you have access to the inference server you are good to go. idea is as such, every n milliseconds of silence (usually 300ms)…

This is the 3rd or so time I've posted this - Run a tiny shitty llm, one that is super fast. Or better yet try out a decision model now days. Feed it the previous conversation turn and all the STT (ASR output) you have so far and just ask it for a probability that the user is done speaking. Use that probability to…

Questions

Is Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech trending on Hacker News right now?

Not at the moment. It was last on the list on Nov 13, 21:22 UTC, after 2 days of trending.

What is the highest rank Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech reached?

#16 on Hacker News · Front page, on Nov 13, 05:13 UTC. Trendoday checks this list every 15 minutes.

How long has Show HN: RealtimeVAD – trading false positives for 110ms earlier end-of-speech been trending?

It was first seen on Nov 12, 2026 and has been on the list on 2 days (73 checks) since.

How much discussion did it get on Hacker News?

26 points and 5 comments; submitted by koljab on Nov 12, 2026.