Nicheloom visits: 58,069 all-time · 7,637 in the last 7 days

The opportunity tracker for new startups.

Today: 1,953 launches · Niche of the day: specialized calculation tools · See what launched today →

RealtimeVAD

trading false positives for 110ms earlier end-of-speech

Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.

Want to reach these readers? Sponsor an issue.

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 matter. If it's wrong and the user continues speaking I can abort the pipeline fast.I benchmarked my solution over ~3200 turns from my assistant logs and it came out ~110 ms faster initially. Still 100 ms now after some a fix which was needed for quiet speech.I made it public if you want to have a look. If you measure it on your data I really appreciate any feedback. There's a benchmark script in the repo.Thx for reading, have a good day folks.

HN · launched Nov 12, 2026 · 32 upvotes (top 22%) · 0 views

Niche: speech processing software · 57 launches, 7 in the last 7 days ▼ -22% · Explore this niche →

Category and tags
  • Horizontal
  • Model & infra
  • Developer
  • AI feature

Visit website ↗ Source listing ↗

More like this: all 57 launches in speech processing software →

Could you build this?

Partial Tuning a voice activity detector for earlier end-of-speech is a signal-processing and ML problem. The pipeline and evaluation harness are vibe-codeable, but good results need labeled speech data and careful tuning.

What it would actually take: A real version requires a labeled speech dataset with precise end-of-speech timestamps, a baseline like Silero VAD, and a novel post-processing or lightweight model. Latency-versus-false-positive evaluation needs audio ML experience.

Discussion

No comments on this launch.

Competitors

Other products that read as similar to this one : 213 launches clear the similarity bar, closest 8 shown.

Attention rank: #61 of 214 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).

Launched 379 days after the earliest competitor.

Related data

The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.