Human AI
Multimodal sensing for wearable AI
At Meta I work within the Human AI team, building the future of human-centric computing. My work centers on interaction — making a device feel natural to use — and understanding — giving it a real sense of the person using it. I focus on the technical direction for sensing across the wearables portfolio: which sensors are needed to deliver those interaction and understanding features.
My fundamental research is on radar. It works without contact or line of sight, is unaffected by lighting, and draws little power, which suits sensing that runs continuously in the background. We have demonstrated face, eye, and body tracking, health monitoring, and speech recognition.
Beyond radar, that spans ultrasound, MEMS, cameras, and lasers. Choosing among them is rarely about which measures best in isolation — it is about what a modality costs in power, space, and compute once it has to share a device with everything else, and still fit something people are willing to wear.
Selected publications: (Ma et al., 2025; Huang et al., 2024; Nguyen et al., 2025)
References
2025
- SenSysmmET: mmWave Radar-Based Eye Tracking on Smart GlassesIn Proceedings of the 23rd ACM Conference on Embedded Networked Sensor Systems, 2025
- BioSensorsWireless Respiration Tracking Using Smart GlassesIn 2025 IEEE BioSensors Conference (BioSensors), 2025
2024
- IEEE MWTLRadar-based heart rate sensing on the smart glassesIEEE Microwave and Wireless Technology Letters, 2024