Research map / Point clouds / Point cloud classification
Improving Graph Representation for Point Cloud Segmentation via Attentive Filtering
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 |
|---|---|---|---|
| Add Spatial Feature Projection to graph convolutions (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 1.1% mIoU | Helped |
| Add graph-shared down-sampling and up-sampling (ablation) | AF-GCN with SFP · S3DIS Area-5 | mIoU: 0.7% mIoU | Helped |
| Add Graph Attentive Filter blocks in the last two stages (ablation) | AF-GCN with SFP and graph-shared sampling · S3DIS Area-5 | mIoU: 71.1 -> 72.3 | Helped |
| Remove Spatial Feature Projection (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 1.9% mIoU | Hurt |
| Remove graph-shared down-sampling and up-sampling (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 0.7% mIoU | Hurt |
| Use GAF only in stage 4 instead of stages 3–4 (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 72.3 -> 71.5 | Hurt |
| Extend GAF blocks into stage 2 (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 72.3 -> 69.3 | Hurt |
| Use GAF blocks at every stage (ablation) | AF-GCN · S3DIS Area-5 | mIoU: 72.3 -> 70.1 | Hurt |
| Scale the AF-GCN model to the large configuration | AF-GCN · S3DIS Area-5 | mIoU: 72.3 -> 73.3 | Helped |
Closest papers by meaning
- Convolution in the Cloud: Learning Deformable Kernels in 3D Graph Convolution Networks for Point Cloud Analysis (CVPR 2020 · similarity 0.90)
- Learning Localized Generative Models for 3D Point Clouds via Graph Convolution (ICLR 2019 · similarity 0.89)
- DHGCN: Dynamic Hop Graph Convolution Network for Self-Supervised Point Cloud Learning (AAAI 2024 · similarity 0.88)
- Taylor Series-Inspired Local Structure Fitting Network for Few-shot Point Cloud Semantic Segmentation (AAAI 2025 · similarity 0.86)
- PointCFormer: A Relation-Based Progressive Feature Extraction Network for Point Cloud Completion (AAAI 2025 · similarity 0.86)
- Generalized Few-shot 3D Point Cloud Segmentation with Vision-Language Model (CVPR 2025 · similarity 0.86)
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.