OmniPhysGS: 3D Constitutive Gaussians for General Physics-Based Dynamics Generation
About
Recently, significant advancements have been made in the reconstruction and generation of 3D assets, including static cases and those with physical interactions. To recover the physical properties of 3D assets, existing methods typically assume that all materials belong to a specific predefined category (e.g., elasticity). However, such assumptions ignore the complex composition of multiple heterogeneous objects in real scenarios and tend to render less physically plausible animation given a wider range of objects. We propose OmniPhysGS for synthesizing a physics-based 3D dynamic scene composed of more general objects. A key design of OmniPhysGS is treating each 3D asset as a collection of constitutive 3D Gaussians. For each Gaussian, its physical material is represented by an ensemble of 12 physical domain-expert sub-models (rubber, metal, honey, water, etc.), which greatly enhances the flexibility of the proposed model. In the implementation, we define a scene by user-specified prompts and supervise the estimation of material weighting factors via a pretrained video diffusion model. Comprehensive experiments demonstrate that OmniPhysGS achieves more general and realistic physical dynamics across a broader spectrum of materials, including elastic, viscoelastic, plastic, and fluid substances, as well as interactions between different materials. Our method surpasses existing methods by approximately 3% to 16% in metrics of visual quality and text alignment.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Observable state simulation | MOSIV Synthetic | PSNR29.79 | 23 | |
| Future state simulation | MOSIV Synthetic | E-P151.8 | 12 | |
| Video Generation | VBench | Image Quality37.8 | 10 | |
| Video Generation | PhysVideo Evaluation 1.0 (test) | Physical Realism Score0.622 | 7 | |
| 4D Physical Generation | GSO | Inference Time12 | 7 | |
| Dynamics Prediction | Phlowers (In-Sequence) | PSNR15.31 | 7 | |
| Dynamics Prediction | Phlowers Cross-Sequence | PSNR15.39 | 7 | |
| Physics-based Dynamic 3DGS Simulation | Physics-based Dynamic 3DGS Simulation Evaluation (test) | CS (Color Similarity)0.265 | 7 | |
| Video Generation | Constructed evaluation dataset | Photo Score58.67 | 7 | |
| Physically-grounded Video Generation | Physically-grounded Video Evaluation Set Human Designed, Real World, and AI Generated scenes | OC17.1 | 5 |