Hi, I’m Guozhen Zhu, the Principal Scientist at Origin AI and ADT Research. I build AI systems that turn raw observations into transferable representations of human state and context — giving machines a reliable sense of what is happening around them. My research centers on foundation-model-based perception and multimodal/temporal learning for ambient intelligence, spanning presence, motion and activity, localization, and health-related sensing. Increasingly, I’m drawn to physical AI: extending this perception into systems that not only understand the world but act in it — helpfully and responsibly.

I care as much about shipping as publishing. At Origin, my research has gone from prototype to products deployed at scale (25M+ routers and 100M+ smart devices worldwide) across consumer and security markets — taking ideas from 0→1 all the way to global rollout. What drives me is turning frontier research into systems people actually use, and building the teams that make that happen. Before Origin, I earned my Ph.D. and M.S. from the ECE SIG at UMD , advised by Dr. K. J. Ray Liu.

Research interests:

• Foundation models for perception: large-scale representation learning over multimodal & temporal data, with strong transfer across tasks and modalities.

• Embodied & physical AI: grounding these representations in LLMs, agents, and robotic systems — reasoning, planning, and control that close the loop from perceiving the world to acting in it.

• Real-world reliability & human-centric intelligence: benchmarks and evaluation for generalization and perceptual limits, in service of understanding human state and context for safety, health, and spatial intelligence.

Outside of work, you’ll often find me on the tennis court , playing the piano , baking something semi-sweet , or soaking in nature’s beauty .

Thanks for stopping by! I’m always open to discussions, new ideas, and meaningful collaborations—whether you’re a researcher, a builder, or exploring what’s next in physical AI—don’t hesitate to reach out!