VAD to the Bone: Ultra-Tiny Speech Activity Detection for Edge Deployment
arXiv:2607.25870
The paper introduces kiloVAD, an ultra‑tiny, CNN‑only voice activity detection model designed for edge devices, using standard Mel features, structured pruning with self‑distillation, and angle‑based quantization‑aware training to achieve state‑of‑the‑art causal performance.
Abstract
Voice activity detection (VAD) triggers downstream speech processing in always-on systems under strict memory, latency, and compute constraints. Recent compact models report strong accuracy but rely on components that are not widely supported: learnable filterbanks, recurrent layers, or non-causal post-processing. We propose kiloVAD, designed for embedded inference using standard Mel features, CNN-only layers, and tunable context/spectral parameters. We introduce per-layer structured pruning with self-distillation and angle-based quantization-aware training (QAT) that outperforms standard QAT by 1-4%. Evaluated per-frame under causal conditions, kiloVAD achieves 0.850 AUC on AVA-Speech with 2.1 k parameters and 200 ms context, establishing a new state of the art for causal, deployment-ready VAD.
Accepted for publication at INTERSPEECH 2026