Home Assistant Integration to install openWakeWord´s from fwartner/home-assistant-wakewords-collection

Train custom wake word models with openWakeWord. A granular 13-step pipeline with compatibility patches for torchaudio 2.10+, Piper TTS, and speechbrain. Generates tiny ONNX models (~200 KB) for real-time keyword detection — like building your own "Hey Siri" trigger. WSL2/Linux + CUDA required.

Offline Vietnamese voice AI assistant built on Nemotron-3.5 ASR. 'Nemo ơi' wake word + local skills + Piper TTS. Same pipeline does real-time translation.

A voice assistant with local LLM as a backend

Custom OpenWakeWord training pipeline for generating wake word models compatible with HAwake Android

Small browser-first wrapper around the OpenWakeWord models using onnxruntime-web. It exposes a WakeWordEngine class you can drop into a React app to listen for wake words like hey_jarvis directly in Chrome, no native layer required.

Trigger-Talk is an offline-capable hotword detection framework that passively listens for custom wake phrases to trigger speech recognition or automation workflows.

Offline AI voice assistant with semantic memory, wake-word detection, local LLM inference, streaming TTS, and modular tool-agent architecture.

A cpp runtime + c shared library for openwakeword detection

Assistente de voz local com detecção de wake word e transcrição de fala que digita o texto direto no campo focado — sem nuvem, sem atalho de teclado.

Local, reproducible openWakeWord custom wake-word training: a working port of the upstream Colab notebook, plus an evaluation harness and the measured results — including the negative ones — of nine model generations.

Angela Anthropomorphized - 3d AI Voice Assistant

Hands-free voice input for Claude Code on Linux

Context-aware voice and command layer for Home Assistant with room-aware, deterministic routing.

Native Go inference pipeline for openWakeWord

Voice satellite for the Hermes agent: openWakeWord + Moonshine STT + Piper TTS on a Raspberry Pi with a ReSpeaker 2-Mic HAT

A local, self-hosted voice assistant with a wake word you set yourself. Runs CPU-only on a Raspberry Pi: wake word, speech-to-text, a persistent-assistant brain, and text-to-speech.

Hands-free voice output for Claude Code on Windows. Free, MIT licensed, no signup, no accounts. Microsoft Edge TTS (cloud) for voices, openWakeWord (local) for wake detection.

Bulletproof openWakeWord trainer Colab notebook (2026 edition). Run-all-and-walk-away. Train your own wake word in 75-90 min on Colab Pro.