Research map / Point clouds / Segmentation and detection
Unsupervised Point Cloud Object Co-Segmentation by Co-Contrastive Learning and Mutual Attention Sampling
Cheng-Kun Yang, Yung-Yu Chuang, Yen-Yu Lin
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 mutual attention instead of self-attention (ablation) | co-contrastive samplers · ScanObjectNN | mIoU: 0.59 -> 0.61 | Improved |
| add mutual attention instead of self-attention (ablation) | co-contrastive samplers · S3DIS | mIoU: 0.45 -> 0.46 | Improved |
| add attention module (mutual) to losses (ablation) | samplers with all three losses · ScanObjectNN | mIoU: 0.57 -> 0.61 | Improved |
| add attention module (mutual) to losses (ablation) | samplers with all three losses · S3DIS | mIoU: 0.43 -> 0.46 | Improved |
| use unsupervised co-segmentation vs cloud-level weak supervision (MPRM) | segmentation · ScanObjectNN OBJ_BG | mIoU: 0.518 -> 0.605 | Improved |
| use our method vs Xu et al. 10% point labels | segmentation · ScanObjectNN OBJ_BG | mIoU: 0.602 vs 0.605 | No clear effect |
| use our method vs AdaCoSeg | segmentation · ScanObjectNN OBJ_BG | mIoU: 0.385 -> 0.605 | Improved |
| use our method vs Xu et al. 10% point labels | segmentation · S3DIS | mIoU: 0.706 vs 0.463 | Hurt |
| vary number of sampled FG/BG points (256/512/1024) (ablation) | samplers · ScanObjectNN OBJ_BG | mIoU: .54 to .61 | No clear effect |
| train BGA with our pseudo labels | PointNet++ classification · ScanObjectNN PB_T50_RS | accuracy (%): 78.5 -> 79.4 | Improved |
| train BGA with our pseudo labels | DGCNN classification · ScanObjectNN PB_T50_RS | accuracy (%): 78.1 -> 79.1 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AdaCoSeg — AdaCoSeg: Adaptive Shape Co-Segmentation With Group Consistency Loss (CVPR 2020) · better on 6 of 6 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
- Attention Discriminant Sampling for Point Clouds (ICCV 2023 · similarity 0.89)
- Rethinking Few-shot 3D Point Cloud Semantic Segmentation (CVPR 2024 · similarity 0.87)
- FreePoint: Unsupervised Point Cloud Instance Segmentation (CVPR 2024 · similarity 0.87)
- Point Cloud Pre-training with Diffusion Models (CVPR 2024 · similarity 0.86)
- Mining and Transferring Feature-Geometry Coherence for Unsupervised Point Cloud Registration (NeurIPS 2024 · similarity 0.86)
- PointCFormer: A Relation-Based Progressive Feature Extraction Network for Point Cloud Completion (AAAI 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.
