Event

Wireless AI for the Internet of Things

We are rapidly moving toward a pervasively connected and sensed world of AIoT (standing for AI + Internet of Things), where objects, machines, humans, and data will be all connected to shape an ecosystem of intelligence. At the core of many IoT applications is not only connectivity but also sensing capability to bridge the physical world and cyberspace.

In this talk, I will present our recent research on wireless AI for seamless, high-precision sensing and tracking for smart homes, healthcare, cars, and beyond. I will first talk about how we unleash the incredible potential of pervasive wireless signals (such as WiFi) for a wide range of sensing applications, including motion detection, vital sign monitoring, fall detection, gait recognition, and gesture control, all in a contactless and sensorless way without cameras or wearables.

I will then present a large-scale decimeter-level indoor tracking system using a single Access Point (AP). Our system covers a large space with a single AP, can be rapidly deployed, and scales to massive clients and buildings with low cost, while maintaining high accuracy. I will conclude the talk with a future research vision centered around building wireless AIoT systems that equip the ubiquitously connected IoT devices around us with advanced sensing intelligence.

Chenshu Wu is an assistant research scientist in the Department of Electrical and Computer Engineering at the University of Maryland.

Registration  (no charge)