Research map / 3D vision / Face and shape
LP-DIF: Learning Local Pattern-Specific Deep Implicit Function for 3D Objects and Scenes
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| cluster local regions by latent-code similarity and assign a pattern-specific decoder to each cluster (ablation) | Local DIF · Burghers | L1 Chamfer Distance: 0.00557 -> 0.00394 | Helped |
| add region re-weighting based on kernel density estimation (ablation) | LP-DIF · Burghers | L1 Chamfer Distance: 0.00419 -> 0.00394 | Helped |
| use dynamic rather than fixed density estimation for region re-weighting (ablation) | LP-DIF · 3D scene reconstruction | geometry detail | Helped |
| replace similarity-based clustering with random assignment of local regions to clusters (ablation) | LP-DIF · 3D shape reconstruction | reconstruction performance | Hurt |
| increase local codes from 32K to 2097K (ablation) | Baseline · Burghers | CD Mean: 0.00557 -> 0.00389 | Helped |
| increase local codes from 32K to 2097K (ablation) | LP-DIF · Burghers | CD Mean: 0.00394 -> 0.00387 | No clear effect |
| use wider 800-dimensional hidden layers to match MDIF's parameter count (ablation) | LP-DIF · ShapeNet airplane | CD: 0.012 -> 0.0145 | Hurt |
Closest papers by meaning
- Local Deep Implicit Functions for 3D Shape (CVPR 2020 · similarity 0.93)
- Learning Deep Implicit Functions for 3D Shapes With Dynamic Code Clouds (CVPR 2022 · similarity 0.93)
- Diffusion 3D Features (Diff3F): Decorating Untextured Shapes with Distilled Semantic Features (CVPR 2024 · similarity 0.87)
- Few-Shot Unsupervised Implicit Neural Shape Representation Learning with Spatial Adversaries (ICML 2024 · similarity 0.87)
- A Lightweight UDF Learning Framework for 3D Reconstruction Based on Local Shape Functions (CVPR 2025 · similarity 0.87)
- Implicit Modeling of Non-rigid Objects with Cross-Category Signals (AAAI 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.