Research map / Point clouds / Shape and surface completion
I-Filtering: Implicit Filtering for Learning Neural Distance Functions From 3D Point Clouds
Shengtao Li, Yudong Liu, Ge Gao, Ming Gu, Yu-Shen Liu
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 zero level set filter (L_zero) (ablation) | L_CD · FAMOUS | Chamfer L2 x1000: 0.031 -> 0.018 | Improved |
| add CD loss to the pull loss (ablation) | L_pull · FAMOUS | Chamfer L2 x1000: 0.083 -> 0.031 | Improved |
| add non-zero level set filtering (L_field) (ablation) | L_CD + L_zero · FAMOUS | Chamfer L2 x1000: 0.018 -> 0.011 | Improved |
| add L_dist term completing the full loss (ablation) | L_CD + L_zero + L_field · FAMOUS | Chamfer L2 x1000: 0.011 -> 0.008 | Improved |
| use bidirectional projection distance instead of single direction (ablation) | implicit filter · FAMOUS | Chamfer L2 x1000: 0.024 -> 0.008 | Improved |
| add Eikonal term (ablation) | full method with CD · FAMOUS | Chamfer L2 x1000: 0.008 -> 0.009 | No clear effect |
| use Eikonal term without CD loss (ablation) | full method · FAMOUS | Chamfer L2 x1000: 0.008 -> 0.021 | Hurt |
| set CD weight alpha_3 to 10 instead of 0 (ablation) | full method · FAMOUS | Chamfer L2 x1000: 0.013 -> 0.008 | Improved |
| raise CD weight alpha_3 from 10 to 100 (ablation) | full method · FAMOUS | Chamfer L2 x1000: 0.008 -> 0.009 | Hurt |
| increase sigma_n from 15 to 30 degrees (ablation) | implicit filter · FAMOUS | Chamfer L2 x1000: 0.008 -> 0.011 | Hurt |
| use diagonal instead of max for sigma_p (ablation) | implicit filter · FAMOUS | Chamfer L2 x1000: 0.008 -> 0.011 | Hurt |
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.
- DIGS — DiGS: Divergence Guided Shape Implicit Neural Representation for Unoriented Point Clouds (CVPR 2022) · better on 7 of 8 shared results
- GridPull — GridPull: Towards Scalability in Learning Implicit Representations from 3D Point Clouds (ICCV 2023) · better on 3 of 3 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
- Implicit Filtering for Learning Neural Signed Distance Functions from 3D Point Clouds (ECCV 2024 · similarity 0.98)
- Neural Unsigned Distance Fields for Implicit Function Learning (NeurIPS 2020 · similarity 0.89)
- Few-Shot Unsupervised Implicit Neural Shape Representation Learning with Spatial Adversaries (ICML 2024 · similarity 0.89)
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.88)
- MultiPull: Detailing Signed Distance Functions by Pulling Multi-Level Queries at Multi-Step (NeurIPS 2024 · similarity 0.87)
- Inferring Neural Signed Distance Functions by Overfitting on Single Noisy Point Clouds through Finetuning Data-Driven based Priors (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.
