Object Detection @ 1 mW: Achieving Always-On Computer Vision at the Edge

Ultra-Low Power Computing Hardware Architecture and Circuits for Artificial Intelligence and Machine-Learning

Batteryless Always-On Wireless Sensing for Full-Stack IoT Insights-as-a-Service

A frame-free event-based approach to low-power real-time machine vision

Enhanced Speech Segregation with Low-Latency DNN processing

STM32Cube.AI: AI productivity boosted on STM32 MCU

16-bit CNN accuracy in 5.5 mm2 package FPGA Human Presence Detection @ 10mW

Energy Efficient, High Performance AI Acceleration from Edge to Cloud

Towards Further Compression of Low-Bitwidth DNNs with Permuted Diagonal Matrices

State-of-the-Art Voice UI Performance on Device