Building a Small Multilingual Offline Voice‑Control Pipeline with ONNX

说明与用途

使用ONNX和Voicute构建了轻量级离线关键词检测管道，适用于IoT设备，实现低延迟、低内存和电池消耗，并且不依赖云服务。

问题

Deploying a lightweight, always‑listening voice‑control system on resource‑constrained devices without relying on continuous cloud transcription.

结果

The KWS pipeline achieved low false‑activation rates, high recall, and acceptable latency on ESP32‑S3, enabling battery‑powered, offline voice control for multilingual commands.

工作流

• Define fixed command set
• Generate initial KWS model (multi‑keyword if needed)
• Test model with real speakers on target microphone
• Deploy model to device and run on‑device inference
• Collect misses and false activations
• Iteratively add hard negatives or real‑voice samples
• Re‑train or fine‑tune model
• Validate final performance on hardware

证据

For this kind of fixed vocabulary, keyword spotting (KWS) can be a better fit than running a general ASR model all the time.

来源

The model is only one part of a usable always‑listening system. A raw score should not automatically trigger a product action.

来源

We measure at least: true‑trigger recall; false activations per hour; distance and speaking volume; different speakers and accents; TV, music, fan and room noise; model size and device memory; end‑to‑end latency and power impact.

来源

For teams without a speech ML group, a practical validation workflow is: define the exact fixed phrases; generate a first model; test it with real speakers before downloading; run it on the target microphone and device; collect misses and false activations; add real‑voice samples or hard negatives when needed.

来源