← Back to all news
arXiv Machine Learning

Flow Mesh: A Velocity-Regularized Flow Matching Framework for Stable Distributional Modeling of Physical Fields on Unstructured Meshes

FlowMesh is a generative framework that produces full-field thermal‑fluid realizations directly on fixed unstructured meshes by combining permutation‑invariant mesh encoding, deterministic latent flow matching, and an l1 penalty on transport velocity. It achieves high probability‑density‑function coverage (91.7%–97.0%) and significant reductions in temperature‑field Kullback–Leibler divergence compared to latent diffusion models, while generating each realization in about 1.2 s—over 500‑fold faster than traditional computational fluid dynamics. The method enables large‑scale, data‑driven characterization of thermal‑fluid states on complex geometries without the need for grid interpolation or stochastic denoising chains.

Related stories

arXiv AI
Aug 19

ChannelFlow-Tools: A Configuration-Driven Pipeline for Generating Machine-Learning-Ready Datasets of 3D Obstructed Channel Flows

ChannelFlow-Tools is an open‑source, configuration‑driven pipeline that generates machine‑learning‑ready datasets for three‑dimensional obstructed channel flows. It combines procedural obstacle geometry generation across six shape families, signed‑distance‑field voxelisation, lattice‑Boltzmann simulation, and packaging into ML‑ready tensors, all driven by reproducible configuration files. The pipeline is validated through mesh‑integrity audits, SDF representation checks, solver benchmarks, and data‑integrity audits, and it has been used to train surrogate models (3D U‑Net, FNO, U‑FNO) that learn geometry‑to‑flow mappings and exhibit physically interpretable behaviour on out‑of‑distribution splits.

By Shubham Kavane, Lukas Schr\"oder, Kajol Kulkarni, Fernando Gonzalez, Harald Koestler
benchmarks
More like this →
Hugging Face Trending Papers
Jun 17

Advances in Scientific Machine Learning for Coupled Fluid Flow and Transport

This chapter reviews recent advances in Scientific Machine Learning (SciML) for modeling coupled fluid flow and transport phenomena governed by the incompressible Navier-Stokes and scalar transport equations. Such systems, found in applications like turbidity currents and thermal convection, feature strong nonlinear coupling and multiscale behavior that make high-fidelity simulations computationally expensive.

benchmarks
More like this →
arXiv Machine Learning
Oct 19

Normalizing Flow Maps

The paper introduces Normalizing Flow Maps (NFM), a new class of exactly invertible pushforward maps that extend normalizing flows and flow maps by learning the integrator of an underlying dynamical system while enabling efficient exact likelihood evaluation. NFMs incorporate two consistency axes—pointwise matching for β1-based regression losses and distributional matching for aligning pushforward distributions with a forward process—allowing scalable training. Experiments on CIFAR-10, ImageNet-64, and Boltzmann Generators demonstrate that NFMs achieve state‑of‑the‑art sample quality among parameter‑matched normalizing flows and outperform prior normalizing flows on peptide benchmarks.

By Francesco M. Ruscio, Timothy James Hitge, Fred Zhangzhi Peng, Danyal Rehman, Zander W. Blasingame, Nicholas M. Boffi, Michael M. Bronstein, Alexander Tong, Avishek Joey Bose
diffusionreinforcement-learningefficiencybenchmarks
More like this →