BIOE Seminar Series: Associate Professor Yuxiong Wang

Imagination with Evidence: AI That Learns More from Less

Abstract: How can AI make more of limited data while keeping its conclusions grounded in evidence? This question connects fundamental challenges in machine learning with opportunities in bioengineering and medicine, where observations can be scarce and meaningful differences subtle. In this talk, I will present our research in computer vision and machine learning along two complementary directions: generating useful training data and reasoning with visual evidence. Using Parkinson’s disease analysis and synthesis of symptom-related facial changes as a motivating example, I will discuss how generative models can support learning with limited supervision and how visual reasoning models can connect their conclusions to relevant observations. I will then explore how these ideas can extend to action through agents that use tools and feedback, with a brief outlook on human motion modeling and robotics. Throughout, I will highlight open challenges and opportunities to combine these AI advances with biomedical expertise in imaging, cell and tissue analysis, biological design, and movement science.

Biography: Yuxiong Wang is an Associate Professor in the Siebel School of Computing and Data Science at the University of Illinois Urbana-Champaign. He is also affiliated with the Artificial Intelligence for Future Agricultural Resilience, Management, and Sustainability (AIFARMS) Institute, the Center for Digital Agriculture (CDA), and the National Center for Supercomputing Applications (NCSA). He received a Ph.D. in robotics from Carnegie Mellon University. His research interests lie in computer vision, machine learning, and robotics, with a particular focus on open-world perception, multimodal learning, generative modeling, and agent learning. He is an ONR Chief of Naval Research Fellow and a recipient of awards including the DARPA Young Faculty Award, the Amazon Faculty Research Award, the ECCV Best Paper Honorable Mention Award, and two CVPR Best Paper Award finalist recognitions. He was also selected to participate in the National Academy of Engineering’s Frontiers of Engineering Symposium. For details: https://yxw.cs.illinois.edu/.