Jianbo Shi

Jianbo Shi is a Professor in the Department of Computer and Information Science at the University of Pennsylvania and a leading researcher in computer vision. His group develops algorithms for understanding images, video, and human behavior at multiple levels of abstraction. Their work spans foundational problems such as object and image segmentation, body pose and limb tracking, gesture and activity recognition, human–object interaction, and video-based inference. Beyond recognition, Shi aims to build computational models of visual thinking—systems that not only perceive their surroundings but also acquire higher-level cognitive abilities such as memory, learning, and structured reasoning. His lab investigates the visual cues that drive human perception and leverages them to design algorithms capable of interpreting complex scenes, predicting actions, and understanding semantic relationships. Shi’s research has influenced areas ranging from medical imaging and augmented reality to video understanding and structured prediction. He also co-founded LYF Innovation Ltd, extending his group's innovations into real-world applications. His long-term vision seeks to unify low-level perception with high-level cognition to enable truly intelligent visual systems.