Research map / Point clouds / Shape and surface completion
Learning Fine-Grained Segmentation of 3D Shapes Without Part Labels
Check 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 |
|---|---|---|---|
| Add low-rank loss to PriorNet (ablation) | PriorNet · Fine-grained shape dataset; categories shown in Table 1 | avg. IoU (%): 55.6 -> 64.1; 68.3 -> 73.8; 72.4 -> 73.0; 72.9 -> 73.8; 75.6 -> 73.3; 65.9 -> 71.9; 78.0 -> 83.1; 35.1 -> 39.3; 42.3 -> | Mixed |
| Use PriorNet+MergeNet instead of PriorNet+BL edge classification | Fine-grained segmentation pipeline · Fine-grained shape dataset; per-category results | avg. IoU (%): 30.7 -> 35.3; 42.4 -> 48.9; 38.7 -> 41.5; 38.7 -> 40.3; 35.1 -> 35.7; 44.9 -> 49.6; 20.5 -> 27.6; 12.9 -> 14.6; 15.8 -> | Helped |
| Use volumetric partition resolution 7^3 rather than the other tested resolutions (ablation) | PriorNet+MergeNet · All categories | average IoU | Helped |
| Use three MergeNet message-passing layers rather than other tested layer counts (ablation) | MergeNet · All categories | average IoU | Mixed |
| Set the low-rank approximation maximum rank too large (ablation) | Low-rank approximation module · Fine-grained shape segmentation | segmentation quality | Hurt |
Closest papers by meaning
- PartNet: A Recursive Part Decomposition Network for Fine-Grained and Hierarchical Shape Segmentation (CVPR 2019 · similarity 0.91)
- PartNet: A Large-Scale Benchmark for Fine-Grained and Hierarchical Part-Level 3D Object Understanding (CVPR 2019 · similarity 0.91)
- SegGraph: Leveraging Graphs of SAM Segments for Few-Shot 3D Part Segmentation (NeurIPS 2025 · similarity 0.87)
- PartNeXt: A Next-Generation Dataset for Fine-Grained and Hierarchical 3D Part Understanding (NeurIPS 2025 · similarity 0.87)
- BFANet: Revisiting 3D Semantic Segmentation with Boundary Feature Analysis (CVPR 2025 · similarity 0.86)
- PartDistill: 3D Shape Part Segmentation by Vision-Language Model Distillation (CVPR 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.