Physics3D: Learning Physical Properties of 3D Gaussians via Video Diffusion

Source

  • Person key: alfredyao
  • Source kind: paper
  • Canonical URL: https://arxiv.org/abs/2406.04338
  • License: NOASSERTION
  • Public handling: public-metadata-summary-hash-link-only
  • Semantic hash: b9cdf81fd45d3ee95c14b5a3ceacc8be4812a80ca89c35c5e9d32cfc04ba1b82
  • First seen: 2026-05-16
  • Last changed: 2026-05-16
  • Identity guard: Do not confuse with yao-shunyu-ysymyth, the OpenAI language-agents researcher at ysymyth.github.io.

Classification

Summary

In recent years, there has been rapid development in 3D generation models, opening up new possibilities for applications such as simulating the dynamic movements of 3D objects and customizing their behaviors. However, current 3D generative models tend to focus only on surface features such as color and shape, neglecting the inherent physical properties that govern the behavior of objects in the real world. To accurately simulate physics-aligned dynamics, it is essential to predict the physical properties of materia…

What This Teaches

How learning, planning, and feedback loops shape agent behavior and model capability.

Public Handling Notes

  • EXTRACTED: Metadata and links are from public sources.
  • INFERRED: Unclear-license full text, PDFs, and source code should not be mirrored into public wiki pages.
  • AMBIGUOUS: Items discovered by keyword search are still separated by source identity and category heuristics.