Computer Vision, Machine Learning, Robotics, Self-supervised Learning, Video Understanding, Common Sense Reasoning
Xiaolong Wang’s research lies in exploiting the structure in data for learning visual representations, with a focus on the spatial-temporal structure in videos and its connection to 3D structure as well as semantic structure. There are two principal directions he has explored: first, to use the structure information from the data itself as a supervisory signal for learning visual representations (i.e., self-supervised learning), eliminating the need for manual labels; second, to explicitly model the structure in data for human activity analysis, scene affordance reasoning and learning object interaction, with potential applications in robotics.
Xiaolong Wang joined UC San Diego as an assistant professor in the Department of Electrical and Computer Engineering in 2020. He obtained his Ph.D. from The Robotics Institute at Carnegie Mellon University. He has collaborated with research labs including Berkeley AI Research, Facebook AI Research, and Allen Institute for Artificial Intelligence. He is the recipient of Facebook Fellowship, Nvidia Fellowship, and Baidu Fellowship.