FIGConvNet
Factorized grids for fast car pressure and drag prediction
Paper title Factorized Implicit Global Convolution for Automotive Computational Fluid Dynamics Prediction
Paper arXiv 2502.04317PDF
Replaces one dense 3D grid around the car with a few implicit grids, each coarse along one axis and fine along the other two. Surface point features are projected onto these grids, mixed by large convolutions rewritten as cheaper 2D ones inside a U-Net and fused across grids; the decoder reads pressure back at the surface points, and the drag coefficient is regressed from the encoder output.
Share on LinkedIn, X or Bluesky, or or .
Know related work that belongs here? Suggest a link
Something wrong on this page? Tell the editor