人妻视频

The goal of the Visual Robotics Laboratory is to build universal embodied intelligent agents. Vision, along with other modalities such as tactile sense and force sense, is an important perceptual modality for embodied intelligence. Our laboratory aims to study how to map these multimodal perceptual inputs to robotic tasks through deep learning. Specifically, our research topics include: control of humanoid robots, dexterous manipulation, VLA models, and tactile sense.