Research map / 3D vision / Face and shape
Neural Vector Fields for Implicit Surface Representation and Inference
Edoardo Mello Rella, Ajad Chhatkuli, Ender Konukoğlu, Luc Van Gool
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 |
|---|---|---|---|
| predict vector field directly instead of unsigned distance (NDF) | Neural Vector Fields · ShapeNet watertight chairs | Chamfer mean: 0.418 -> 0.193 | Improved |
| predict vector field directly instead of NDF | Neural Vector Fields · MGN clothes (open) | Chamfer mean: 0.786 -> 0.128 | Improved |
| use VF instead of SDF-based DeepSDF | Neural Vector Fields · ShapeNet watertight sofas | Chamfer mean: 0.120 -> 0.152 | Hurt |
| use VF instead of DeepSDF | Neural Vector Fields · ShapeNet watertight chairs | F1-Score: 89.31 -> 87.57 | Hurt |
| apply marching cubes on NDF distance instead of directly predicting directions | NDF · ShapeNet watertight chairs | Chamfer mean: 0.418 -> 0.629 | Hurt |
| add piecewise planar prior (Planar VF) (ablation) | VF · ShapeNet planar bookshelves | Chamfer mean: 1.116 -> 0.558 | Improved |
| add piecewise planar prior (Planar VF) (ablation) | VF · ShapeNet planar laptops | F1-Score: 94.38 -> 95.76 | Improved |
| use larger backbone network (ablation) | VF · ShapeNet cars (larger backbone) | Chamfer mean: 0.0122 vs NDF 0.0126 | Improved |
| compare VF to differentiated UDF for normals | VF vs UDF · Normal consistency (cube toy) | L1 distance: 0.149 -> 0.125 | Improved |
| use VF instead of UDF gradients on tetrahedron shape | VF vs UDF · Double tetrahedron | L1 error: 0.4862 -> 0.1831 | Improved |
| use VF instead of UDF gradients on stress test | VF vs UDF · Complex tetrahedron stress test | L1 error: 0.7952 -> 0.3085 | Improved |
| use VF instead of SDF/UDF for normal consistency | VF · ShapeNet chairs normals | cosine similarity: 0.874 -> 0.898 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- GIFS — GIFS: Neural Implicit Function for General Shape Representation (CVPR 2022) · better on 9 of 9 shared results
- DeepSDF — DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation (CVPR 2019) · better on 4 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Neural Vector Fields: Implicit Representation by Explicit Learning (CVPR 2023 · similarity 0.94)
- Neural Vector Fields: Generalizing Distance Vector Fields by Codebooks and Zero-Curl Regularization (TPAMI 2025 · similarity 0.92)
- Unsigned Orthogonal Distance Fields: An Accurate Neural Implicit Representation for Diverse 3D Shapes (CVPR 2024 · similarity 0.88)
- An Intrinsic Vector Heat Network (ICML 2024 · similarity 0.88)
- DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling (CVPR 2024 · similarity 0.87)
- CoFie: Learning Compact Neural Surface Representations with Coordinate Fields (NeurIPS 2024 · similarity 0.87)
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.
