Xiaoyang Wu (吴笑阳)

What if an intelligent system could know things about you — your body, your environment, the people around you — that you cannot fully articulate yourself? What if AI could model all of it across modalities, and respond not only in language, but across the full range of human sensation? What if humans and intelligent systems could become extensions of each other's perception — a closed loop where AI senses what humans cannot, and humans ground what AI cannot yet feel?

I build ubiquitous sensing systems that capture the continuous, implicit signals humans and their environments generate — physiological, acoustic, tactile, and beyond — expanding the sensing infrastructure available to intelligent systems. I develop multimodal AI that reasons and generates across the full heterogeneity of this signal space, bridging modalities as different as biosignals and chemosensory experience, enabling machines to model and express aspects of human life that language alone cannot reach. I build embodied, multisensory interaction systems that close this loop — where AI and humans communicate through physical and sensory channels, each continuously shaping the other's perception and understanding.

My long-term vision is a future where the richness of human experience — sensory, physiological, social — becomes the native language of intelligent systems, and human flourishing expands with it.