Mesh2NeRF: Direct Mesh Supervision for Neural Radiance Field Representation and Generation
This addresses the issue of artifacts in 3D generative models for computer vision researchers, offering an incremental improvement by providing more accurate supervision for NeRF training.
The paper tackles the problem of generating ground-truth radiance fields from 3D meshes for training neural radiance fields (NeRFs), proposing Mesh2NeRF to directly derive these fields analytically, which results in a 3.12dB PSNR improvement for view synthesis on the ABO dataset and a 0.69 PSNR enhancement in single-view conditional generation on ShapeNet Cars.
We present Mesh2NeRF, an approach to derive ground-truth radiance fields from textured meshes for 3D generation tasks. Many 3D generative approaches represent 3D scenes as radiance fields for training. Their ground-truth radiance fields are usually fitted from multi-view renderings from a large-scale synthetic 3D dataset, which often results in artifacts due to occlusions or under-fitting issues. In Mesh2NeRF, we propose an analytic solution to directly obtain ground-truth radiance fields from 3D meshes, characterizing the density field with an occupancy function featuring a defined surface thickness, and determining view-dependent color through a reflection function considering both the mesh and environment lighting. Mesh2NeRF extracts accurate radiance fields which provides direct supervision for training generative NeRFs and single scene representation. We validate the effectiveness of Mesh2NeRF across various tasks, achieving a noteworthy 3.12dB improvement in PSNR for view synthesis in single scene representation on the ABO dataset, a 0.69 PSNR enhancement in the single-view conditional generation of ShapeNet Cars, and notably improved mesh extraction from NeRF in the unconditional generation of Objaverse Mugs.