Seeing the Wind from a Falling Leaf

Jiajun Wu (Stanford University) · Yue Wang (Soochow University, China) · Zhiyuan Gao (University of Southern California) · Jiageng Mao (University of Southern California) · Hong-Xing "Koven" Yu (Stanford University) · HAOZHE LOU (University of Southern California) · Emily Jia (University of Southern California) · Jernej Barbic (University of Southern California)
backpropagationdifferentiable frameworkforce fieldsforce representationsinverse graphicsinvisible forcesmotion modelingobject geometryphysical interactionsphysical propertiesphysics-based generationreal-world scenariossynthetic scenariosvideo editingvisual observations

A longstanding goal in computer vision is to model motions from videos, while the representations behind motions, i.e. the invisible physical interactions that cause objects to deform and move, remain largely unexplored. In this paper, we study how to recover the invisible forces from visual observations, e.g., estimating the wind field by observing a leaf falling to the ground. Our key innovation is an end-to-end differentiable inverse graphics framework, which jointly models object geometry, physical properties, and interactions directly from videos. Through backpropagation, our approach enables the recovery of force representations from object motions. We validate our method on both synthetic and real-world scenarios, and the results demonstrate its ability to infer plausible force fields from videos. Furthermore, we show the potential applications of our approach, including physics-based video generation and editing. We hope our approach sheds light on understanding and modeling the physical process behind pixels, bridging the gap between vision and physics. Please check more video results in our project page https://chaoren2357.github.io/seeingthewind/ .