ESP32 voice assistant: 8x wake-word model compression with AIMET
carrycooldude · x · 2026-09-01
Ashwin shared a project building a small voice assistant for ESP32. The wake-word model was compressed 8x using Qualcomm's AIMET. While naive 4-bit rounding caused a 3.7% accuracy loss, AIMET's AdaRound successfully recovered the performance. The wake words were generated by GradiumAI TTS in 22 different voices.
More from Embodied
- Turing Institute AI model monitors satellites to predict orbital collisions — turinginst · 2026-09-01
- Discussion sparks around Nanoduck: a pocket-sized desk robot concept — RachelVT42 · 2026-09-01
- Embodiment-aware RL for robot control presented at three international workshops — Jan_R_Peters · 2026-09-01
- XPeng Humanoid Robot HQ Walkthrough: Gait and Confidence Need Improvement — TinfoilTricorn · 2026-09-01
- AeroVect Shows Real-Time Robot Perception System for Airport Ramps — lukas_m_ziegler · 2026-09-01
- LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigation — LightOriginsHQ · 2026-09-01