paper-with-me

Papers

Pretrained Video Models as Differentiable Physics Simulators for Urban Wind Flows

2026-03-22 · Janne Perini, Rafael Bischof, Moab Arar, Ayça Duran, Michael A. Kraus, Siddhartha Mishra, Bernd Bickel arxiv

Designing urban spaces that provide pedestrian wind comfort and safety requires time-resolved Computational Fluid Dynamics (CFD) simulations, but their current computational cost makes extensive design exploration impractical. We introduce WinDiNet (Wind Diffusion Network), a pretrained video diffusion model that is repurposed as a fast, differentiable surrogate for this task. Starting from LTX-Video, a 2B-parameter latent video transformer, we fine-tune on 10,000 2D incompressible CFD simulations over procedurally generated building layouts. A systematic study of training regimes, conditioning mechanisms, and VAE adaptation strategies, including a physics-informed decoder loss, identifies a configuration that outperforms purpose-built neural PDE solvers. The resulting model generates full 112-frame rollouts in under a second. As the surrogate is end-to-end differentiable, it doubles as a physics simulator for gradient-based inverse optimization: given an urban footprint layout, we optimize building positions directly through backpropagation to improve wind safety as well as pedestrian wind comfort. Experiments on single- and multi-inlet layouts show that the optimizer discovers effective layouts even under challenging multi-objective configurations, with all improvements confirmed by ground-truth CFD simulations.

📄 PDF Abstract BibTeX arXiv:2603.21210

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Learned Neural Physics Simulation for Articulated 3D Human Pose Reconstruction

2024-10-15 · Mykhaylo Andriluka, Baruch Tabanpour, C. Daniel Freeman, Cristian Sminchisescu

We propose a novel neural network approach, LARP (Learned Articulated Rigid body Physics), to model the dynamics of articulated human motion with contact. Our goal is to develop a faster and more convenient methodologica…

Leveraging Reward Gradients For Reinforcement Learning in Differentiable Physics Simulations

2022-03-06 · Sean Gillen, Katie Byl

In recent years, fully differentiable rigid body physics simulators have been developed, which can be used to simulate a wide range of robotic systems. In the context of reinforcement learning for control, these simulato…

Deep Reinforcement Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Differentiability in Unrolled Training of Neural Physics Simulators on Transient Dynamics

2024-02-20 · Bjoern List, Li-Wei Chen, Kartik Bali, Nils Thuerey

Unrolling training trajectories over time strongly influences the inference accuracy of neural network-augmented physics simulators. We analyze this in three variants of training neural time-steppers. In addition to one-…

Differentiable Physics Simulations with Contacts: Do They Have Correct Gradients w.r.t. Position, Velocity and Control?

2022-07-08 · Yaofeng Desmond Zhong, Jiequn Han, Georgia Olympia Brikis

In recent years, an increasing amount of work has focused on differentiable physics simulation and has produced a set of open source projects such as Tiny Differentiable Simulator, Nimble Physics, diffTaichi, Brax, Warp,…

Position

Radio Propagation Modelling: To Differentiate or To Deep Learn, That Is The Question

2025-09-15 · Stefanos Bakirtzis, Paul Almasan, José Suárez-Varela, Gabriel O. Ferreira 외 arxiv

Differentiable ray tracing has recently challenged the status quo in radio propagation modelling and digital twinning. Promising unprecedented speed and the ability to learn from real-world data, it offers a real alterna…