Xi Wang

Dr. Xi Wang is a postdoctoral researcher at the Technical University of Munich (TUM) within the Computer Vision Group (Informatics 9) under Prof. Daniel Cremers, and a researcher at the Computer Vision and Geometry Lab at ETH Zurich led by Prof. Marc Pollefeys. He collaborates with Prof. Luc Van Gool at INSAIT and has affiliations with MIT and Adobe Research. His research spans computer vision, computer graphics, vision science, and vision-language multimodal learning, focusing on modeling human common sense, intent-driven behavior, and egocentric scene understanding.

Research Highlights:

• Developed multimodal world models (GEM) for fine-grained ego-motion and object dynamics
• Created vision-language interior design systems (I-Design) and driver trajectory prediction models
• Advanced 3D human-object interaction reconstruction (ROMEO) and eye-tracking-based action anticipation
• Explored Gaussian splatting for dynamic scene understanding (EgoGaussian)

Scientific Recognition:

• Recipient of the ECCV 2024 Workshop Best Paper Award for ROMEO
• $1.6 million grant from BMBF for establishing his independent TUM research group
• Accepted papers at top venues: CVPR, ECCV, ICLR, ETRA, ISMAR