01 / PERCEPTION
Video and Visual Understanding
Developing models that understand objects, scenes, actions, and temporal structure from images and videos.
Computer vision · Video understandingWhat we explore
Research Projects
Our research brings together perception, learning, reasoning, and control for intelligent systems operating in dynamic environments.
02 / MULTIMODAL LEARNING
Multimodal Learning
Connecting visual, language, and other data modalities to build representations that support robust understanding and reasoning.
Vision-language models · Representation learning03 / CONTROL
Learning for Control
Studying how intelligent systems can make decisions and act reliably under changing conditions.
World modeling · Vision-Language-Action · Perception and action