Neural method adds detail to specific mesh areas without changing rest
Learned Localized Mesh Refinement
Computational Geometry
Summary
Meshes are like 3D shapes made of triangles but making them more detailed everywhere can be slow and waste resources. The authors created a neural network that can add details only to parts of a mesh people want to improve, keeping the rest the same. Their approach learns many ways to refine small regions while respecting the whole shape. This selective refining helps add detail where it matters most, unlike usual methods that refine everything or none. This method can help create better 3D models more efficiently for things like video games or simulations.
What this means in practice
- •For 3d graphics developers: Refine specific regions of 3D models during rendering to optimize detail and performance based on viewer focus.
- •For simulation engineers: Adjust mesh detail adaptively in physically important regions to improve simulation accuracy without increasing mesh size globally.
Authors
Xiao Zhan, Chrystiano Araújo, Kangle Deng, Maneesh Agrawala, Hsueh-Ti Derek Liu, Mina Konaković Luković
Abstract
We present a neural method for adaptive triangle mesh refinement, in which an autoregressive model adds geometric detail to selected regions of an input mesh while leaving the rest unchanged, a key capability for efficiently allocating mesh budget. Existing upsampling methods struggle to achieve this. Classical subdivision schemes refine triangulation without semantic awareness of the underlying shape or the ability to recover geometric details missing from a coarse input. Recent neural mesh models generate shapes globally, sacrificing region-specific control. We propose a novel tokenizer that yields combinatorially many valid upsampling trajectories from a single mesh. Trained on such data, our locally-conditioned autoregressive architecture allows for direct manipulation of topology and geometry within target regions of an input mesh. We validate our method against state-of-the-art approaches and demonstrate its ability to perform adaptive upsampling with region-selective control, a capability absent from existing approaches. This unlocks inference-time view-dependent refinement, physics-aware region refinement, and coarse-shape conditioned novel mesh synthesis. We provide code at https://github.com/seanxzhan/learned-localized-mesh-refinement/.