Research map / 3D vision / Face and shape
UNIS: A Unified Framework for Achieving Unbiased Neural Implicit Surfaces in Volume Rendering
Junkai Deng, Hanting Niu, Jiaze Li, Fei Hou, Ying He
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 |
|---|---|---|---|
| replace VolSDF mapping with unbiased UNIS-C1 mapping | VolSDF · DTU (no mask supervision) | Chamfer distance mean: 0.857 -> 0.778 | Improved |
| UNIS-C3 versus NeuS | NeuS · DTU (no mask supervision) | Chamfer distance mean: 0.837 -> 0.777 | Improved |
| UNIS-C1 versus VolSDF | VolSDF · OmniObject3D (no mask supervision) | Chamfer distance mean: 3.277 -> 3.192 | Improved |
| UNIS-C1 versus NeuS | NeuS · OmniObject3D (no mask supervision) | Chamfer distance mean: 3.094 -> 3.192 | Hurt |
| UNIS-C4 versus VolSDF | VolSDF · OmniObject3D (no mask supervision) | Chamfer distance mean: 3.277 -> 4.037 | Hurt |
| UNIS-C5 versus NeuS on DTU | NeuS · DTU (no mask supervision) | Chamfer distance mean: 0.837 -> 0.895 | Hurt |
| use SDF instead of GWN implicit function (k=1, logistic) (ablation) | UNIS · OmniObject3D Antique | Chamfer distance: 2.81 -> 1.81 | Improved |
| use SDF instead of GWN implicit function (k=1, Laplace) (ablation) | UNIS · OmniObject3D Antique | Chamfer distance: 2.58 -> 1.97 | Improved |
| use SDF instead of GWN implicit function (k=2, logistic) (ablation) | UNIS · OmniObject3D Antique | Chamfer distance: 8.23 -> 22.05 | Hurt |
| increase exponent k from 1 to 2 (logistic, SDF) (ablation) | UNIS · OmniObject3D Antique | Chamfer distance: 1.81 -> 22.05 | Hurt |
| increase exponent k from 1 to 2 (logistic, SDF) on PSNR (ablation) | UNIS · OmniObject3D Antique | PSNR: 36.0636 -> 24.6451 | Hurt |
| use Laplace instead of logistic kernel (k=1, SDF) (ablation) | UNIS · OmniObject3D Antique | Chamfer distance: 1.81 -> 1.97 | Hurt |
Closest papers by meaning
- UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction (ICCV 2021 · similarity 0.92)
- Towards Unbiased Volume Rendering of Neural Implicit Surfaces With Geometry Priors (CVPR 2023 · similarity 0.92)
- Geometry-Aware Projective Mapping for Unbounded Neural Radiance Fields (ICLR 2024 · similarity 0.90)
- NeAT: Learning Neural Implicit Surfaces With Arbitrary Topologies From Multi-View Images (CVPR 2023 · similarity 0.90)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.88)
- Differentiable Neural Surface Refinement for Modeling Transparent Objects (CVPR 2024 · similarity 0.88)
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.
