Research map / 3D vision / Face and shape
DR-KFS: A Differentiable Visual Similarity Metric for 3D Shape Reconstruction
Jiongchao Jin, Akshay Gadi Patil, Zhang Xiong, Hao Zhang
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 |
|---|---|---|---|
| train with DR-KFS instead of original loss | AtlasNet25 · ShapeNet 4 categories (avg) | LFD: 4338 -> 3796 | Improved |
| train with DR-KFS instead of original loss | Matryoshka Net · ShapeNet 4 categories (avg) | LFD: 3866 -> 3675 | Improved |
| train with DR-KFS instead of original loss | Pixel2Mesh · ShapeNet 4 categories (avg) | LFD: 4508 -> 4212 | Improved |
| train with DR-KFS instead of original loss | OGN · ShapeNet 4 categories (avg) | LFD: 4436 -> 4237 | Improved |
| train with DR-KFS instead of original loss | 3D-R2N2 · ShapeNet 4 categories (avg) | LFD: 4692 -> 4416 | Improved |
| train with DR-KFS instead of original loss | AtlasNet25 · ShapeNet 4 categories (avg) | Chamfer distance: 6.53 -> 7.11 | Hurt |
| train with DR-KFS instead of original loss | Pixel2Mesh · ShapeNet 4 categories (avg) | Chamfer distance: 6.61 -> 6.98 | Hurt |
| train with DR-KFS instead of original loss | OGN · ShapeNet 4 categories (avg) | Chamfer distance: 6.13 -> 6.08 | No clear effect |
| train with DR-KFS instead of original loss | Matryoshka Net · ShapeNet 4 categories (avg) | IoU: 65.33 -> 66.08 | Improved |
| train with DR-KFS instead of original loss | 3D-R2N2 · ShapeNet 4 categories (avg) | IoU: 53.98 -> 52.74 | Hurt |
| train with DR-KFS instead of original loss | all tested networks · ShapeNet 4 categories (avg) | visual similarity metrics: 24 out of 30 | Improved |
| train with DR-KFS instead of original loss | all tested networks · ShapeNet 4 categories (avg) | CD and IoU: 4 out of 10 | Mixed |
Closest papers by meaning
- SDFDiff: Differentiable Rendering of Signed Distance Fields for 3D Shape Optimization (CVPR 2020 · similarity 0.87)
- Hard Example Generation by Texture Synthesis for Cross-domain Shape Similarity Learning (NeurIPS 2020 · similarity 0.87)
- SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images (NeurIPS 2020 · similarity 0.86)
- Learning a Multi-View Stereo Machine (NeurIPS 2017 · similarity 0.86)
- DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling (CVPR 2024 · similarity 0.85)
- Diffeomorphic Mesh Deformation via Efficient Optimal Transport for Cortical Surface Reconstruction (ICLR 2024 · 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.
