PhysGen: Rigid-Body Physics-Grounded Image-to-Video Generation

TL;DR

PhysGen introduces a physics-grounded image-to-video generation framework that combines physical simulation with generative models to produce realistic, controllable, and physically plausible videos from a single image and user input.

Contribution

The paper presents a novel integration of model-based physics simulation with data-driven video generation for improved realism and control in image-to-video synthesis.

Findings

PhysGen outperforms existing methods in realism and physical plausibility.

The system enables user-controlled dynamic interactions in generated videos.

Quantitative and user studies confirm superior performance over prior approaches.

Abstract

We present PhysGen, a novel image-to-video generation method that converts a single image and an input condition (e.g., force and torque applied to an object in the image) to produce a realistic, physically plausible, and temporally consistent video. Our key insight is to integrate model-based physical simulation with a data-driven video generation process, enabling plausible image-space dynamics. At the heart of our system are three core components: (i) an image understanding module that effectively captures the geometry, materials, and physical parameters of the image; (ii) an image-space dynamics simulation model that utilizes rigid-body physics and inferred parameters to simulate realistic behaviors; and (iii) an image-based rendering and refinement module that leverages generative video diffusion to produce realistic video footage featuring the simulated motion. The resulting videos are realistic in both physics and appearance and are even precisely controllable, showcasing superior results over existing data-driven image-to-video generation works through quantitative comparison and comprehensive user study. PhysGen's resulting videos can be used for various downstream applications, such as turning an image into a realistic animation or allowing users to interact with the image and create various dynamics. Project page: https://stevenlsw.github.io/physgen/