The National Science Foundation (NSF) awarded a $240,000 Project Grant under the Engineering program (CFDA 47.041) to the University of Illinois Chicago (UIC) to develop novel 3D environment perception capabilities. The project aims to address real-time environment perception using only 2D annotations during training, by modeling and learning the rich space-time structures of the physical environment. Key activities include: (1) establishing a self-supervised framework for lifting 2D objects to 3D through integrating geometry, motion, and differentiable depth rendering, (2) scaling 2D-to-3D perception to the entire environment by modeling short-term and long-term dynamics, and (3) achieving flexible and generalizable environment perception by handling articulated motion and diverse real-world settings. The project period is from August 1, 2024 to July 31, 2027. This research is expected to significantly impact cyber-physical systems and enable a wide range of applications, from robotics to autonomous vehicles.