# XPandora/PhysGaussian

[CVPR 2024 Highlight] PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics

Repository: https://github.com/XPandora/PhysGaussian
Canonical: https://ross.abutalabs.com/products/physgaussian
Language: Python
License Family: other
Last push: 2026-01-21T09:52:09+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 63, release rhythm 35, longevity 73
- inputs: {"age_days": 1024, "days_push": 224, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1414, forks 72 (observed 2026-08-28T04:04:39.618903+00:00)

## What it is
PhysGaussian is a research library that integrates Material Point Method (MPM) physics simulation with 3D Gaussian Splatting representations to generate physically plausible novel motion and dynamics. It evolves the same 3D Gaussian kernels used for rendering through continuum mechanics, avoiding meshing or geometry embeddings.

## Use cases
- simulate elastic objects from gaussian splatting reconstructions
- generate novel motion of 3d scenes with physics
- render deformable materials like fluids and metals from photos
- animate a 3d gaussian splatting model with continuum mechanics
- create dynamic 3d content without meshing
- research physics-integrated neural rendering

## When to choose
- you need physically grounded animation of 3D Gaussian Splatting scenes
- you want to simulate diverse materials (elastic, plastic, fluids, granular) directly on Gaussian kernels
- you are doing research in physics-integrated neural rendering

## When to avoid
- you need a production-ready tool with a stable API or license
- you want real-time physics simulation on consumer hardware
- you need simple static 3D reconstruction without dynamics

## Facets
- artifact type: library
- maturity: active
- function: simulation, graphics, machine-learning, computer-vision
- domain: computer-vision, graphics, simulation, machine-learning
- platform: python
- tags: gaussian-splatting, material-point-method, physics-simulation, 3d-reconstruction, generative-dynamics, research-code, cvpr-2024, linux, gpu

## Member repositories
- XPandora/PhysGaussian (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:39.618903+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:38:11.879785+00:00, confidence not recorded.
  - readme: https://github.com/XPandora/PhysGaussian (fetched 2026-08-28T04:04:39.618903+00:00, sha 6a9e959af4b1)
- Data as of 2026-08-30T08:39:29.467469+00:00.