In this work, we propose MeshFeat, a parametric feature encoding tailored to meshes, for which we adapt the idea of multi-resolution feature grids from Euclidean space.

In this work, we propose MeshFeat, a parametric feature encoding tailored to meshes, for which we adapt the idea of multi-resolution feature grids from Euclidean space.

MeshFeat: Multi-Resolution Features for Neural Fields on Meshes (M Mahajan, F Hofherr and D Cremers), In European Conference on Computer Vision (ECCV), 2024. ( [project page])

Jul 18, 2024 · In this work, we propose MeshFeat, a parametric feature encoding tailored to meshes, for which we adapt the idea of multi-resolution feature grids from Euclidean space.

Aug 28, 2025 · In this work, we introduce a novel geometry-aware framework for representing neural fields on triangle meshes that are multi-resolution across both spatial and frequency domains.

Video for MeshFeat: Multi-Resolution Features for Neural Fields on Meshes. Work done by Mihir Mahajan, Florian Hofherr and Daniel Cremers.

MeshFeat Public MeshFeat: Multi-Resolution Features for Neural Fields on Meshes Python 5