ActingAI: Embodied Interaction for Perception and Learning

The goal of ActingAI-Lab is to activate AI by embodied intelligence, which bridges the gap between perception, learning and action. We focus on some important aspects of embodied intelligence research, including multi-modal active perception, cross-modal lifelong learning, agent-environment dynamic interaction, and multi-agent perception and learning. They can have extensive applications for robotics, industrial intelligence, visually-impaired assistance, etc. Besides, we attempt to establish cloud-edge collaboration architectures for numerous embodied perception and learning applications.

We are affiliated with the Tsinghua University department of Computer Science and Technology, as well as Beijing National Research Center for Information Science and Technology.