PhysLDM: Latent Diffusion for High-Fidelity Deformable Simulation

arXiv:2610.07609 · cs.CV, cs.GR · Submitted 2026-10-06 · Read on arXiv

cs.CV, cs.GR

Submitted: 2026-10-06

Updated: 2026-10-08

Code: https://github.com/newton-physics/newton

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Fidelity — 1× in this paper · explained in 2 episodes

Sources

• PhysiFormer: Learning to Simulate Mechanics in World Space
• RigidFormer: Learning Rigid Dynamics using Transformers
• NeuROK: Generative 4D Neural Object Kinematics
• Axial Attention in Multidimensional Transformers
• WorldParticle: Unified World Simulation of Lagrangian Particle Dynamics via Transformer
• ClothTransformer: Unified Latent-Space Transformers for Scalable Cloth Simulation

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models