Robotics
Robotics foundation models for physical AI—VLAs, dexterous manipulation, and learning from human and robot experience.
Research
Space AI
Reliable Understanding and Reasoning for Earth Observation
3D Vision
Spatial intelligence for the physical world—reconstruction, understanding, generation, and embodied interaction.
Egocentric Vision
Intelligence from the First-Person Perspective
Visual Media
Processing, Editing, and Generating the Visual World
Vision Agent
From Multimodal Foundation Models to Executable Visual Intelligence