jxlarrea

jxlarrea / wyoming-voice-match

Public

A Wyoming protocol ASR proxy that verifies speaker identity and isolates voice commands from background noise before forwarding audio to a downstream speech-to-text service. Designed for Home Assistant voice pipelines to prevent false activations from TVs, radios, and other people - and to deliver clean transcripts even in noisy environments.

asr asr-services docker embeddings home-assistant
31
1
100% credibility
Found Feb 17, 2026 at 21 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A voice verification and speaker extraction service for Home Assistant that ensures only authorized users' clean audio reaches speech-to-text services by filtering background noise and other speakers.

How It Works

1
😩 Voice Assistant Hears TV Noise

Your smart home voice helper picks up TV chatter or others talking, causing wrong actions or messy results.

2
🔍 Discover Voice Match Helper

You find this handy tool that checks if it's really you speaking and cleans up the audio before sending it on.

3
🎙️ Record Your Voice Samples

Speak natural commands into your microphone a bunch of times at different volumes and distances to capture your unique voice.

4
🧠 Train It to Know You

The tool listens to your recordings and learns your voice pattern, so it can spot you anywhere.

5
🔗 Connect to Your Smart Home

Link this voice checker into your home assistant setup, pointing it to handle audio after the wake word.

6
🧪 Test with Background Noise

Say commands with TV blasting and watch it ignore the noise, only passing your clean voice forward.

🎉 Flawless Voice Commands

Now your smart home only responds to you, delivers crystal-clear instructions, and ignores distractions perfectly.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 21 to 31 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is wyoming-voice-match?

Wyoming-voice-match is a Python-based Wyoming protocol ASR proxy that slots into Home Assistant voice pipelines, verifying your speaker identity and stripping background noise like TV chatter or other voices before forwarding clean audio to services such as github wyoming-faster-whisper or github wyoming-whisper. After a wake word from a github wyoming-openwakeword satellite, it buffers the stream, runs quick voiceprint checks, and delivers isolated commands to prevent false activations and garbled transcripts. Deploy via wyoming protocol docker with simple docker-compose.yml pointing to your wyoming protocol host and port, like 10350.

Why is it gaining traction?

It tackles real-world Home Assistant pain points—rejecting TV-induced activations and other speakers while extracting your voice for crisp ASR results in noisy homes—without fixed audio limits or high latency (5-25ms on GPU). Tune thresholds via env vars for your setup, enroll voices with CLI scripts recording from satellites or mics, and get speaker tagging for LLM assistants. Docker images for CPU/GPU make it a drop-in for wyoming protocol server chains with github wyoming-piper TTS.

Who should use this?

Home Assistant users running voice satellites in living rooms with TVs or radios, seeking reliable ASR amid background audio. Multi-person households wanting per-speaker verification to avoid cross-talk activations. Devs building wyoming protocol home assistant host port integrations for robust, noise-resistant assistants.

Verdict

Solid early pick for noisy Home Assistant setups—excellent docs and scripts make enrollment straightforward despite 21 stars and 1.0% credibility score signaling low maturity. Test it if false activations plague your pipeline; skip for quiet environments or production without more community validation.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.