Research map / 3D vision / Face and shape
Intrinsic Neural Fields: Learning Functions on Manifolds
Lukas Koestler, Daniel Grittner, Michael Moeller, Daniel Cremers, Zorah Lähner
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add more eigenfunctions (1023 instead of 64) (ablation) | Intrinsic neural field · Texture reconstruction cat | PSNR: 31.18 -> 34.82 | Improved |
| add more eigenfunctions (1023 instead of 4) (ablation) | Intrinsic neural field · Texture reconstruction cat | PSNR: 22.84 -> 34.82 | Improved |
| use the larger MLP (329k instead of 140k params) (ablation) | Intrinsic neural field · Texture reconstruction cat | PSNR: 34.57 -> 34.82 | Improved |
| use the larger MLP (329k instead of 140k params) (ablation) | Intrinsic neural field · Texture reconstruction human | LPIPS: 0.379 -> 0.306 | Improved |
| use intrinsic LBO embedding instead of random Fourier features (sigma=8) | Texture Fields · Texture reconstruction cat | PSNR: 34.39 -> 34.82 | Improved |
| use intrinsic LBO embedding instead of random Fourier features (sigma=8), 200 epochs | Texture Fields · Texture reconstruction human | PSNR: 31.85 -> 32.37 | Improved |
| use intrinsic LBO embedding instead of mapping to a sphere | NeuTex · Texture reconstruction cat | PSNR: 31.96 -> 34.82 | Improved |
| increase NeuTex embedding size | NeuTex · Texture reconstruction human | PSNR: 29.49 -> 29.22 | Mixed |
| change RFF sigma from 8 to 16 | TF+RFF · Texture reconstruction human | PSNR: 32.26 -> 31.53 | Hurt |
| train for 1000 instead of 200 epochs | Intrinsic neural field · Texture reconstruction cat | PSNR: 34.79 -> 34.82 | No clear effect |
| sample point cloud with fewer points than vertices instead of mesh | Intrinsic neural field · Discretization robustness cat | PSNR: 34.82 -> 33.17 | Hurt |
| remesh with uniform isotropic remeshing instead of original mesh | Intrinsic neural field · Discretization robustness human | PSNR: 32.48 -> 32.63 | No clear effect |
Closest papers by meaning
- Grounding Continuous Representations in Geometry: Equivariant Neural Fields (ICLR 2025 · similarity 0.87)
- Neural Fields as Distributions: Signal Processing Beyond Euclidean Space (CVPR 2024 · similarity 0.87)
- Unsigned Orthogonal Distance Fields: An Accurate Neural Implicit Representation for Diverse 3D Shapes (CVPR 2024 · similarity 0.86)
- How to Train Neural Field Representations: A Comprehensive Study and Benchmark (CVPR 2024 · similarity 0.86)
- CoFie: Learning Compact Neural Surface Representations with Coordinate Fields (NeurIPS 2024 · similarity 0.86)
- Building Transformation Layers for Riemannian Neural Networks (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
