Research map / Point clouds / Segmentation and detection
Salient Object Detection for Point Clouds
Songlin Fan, Wei Gao, Ge Li
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 SPB to PointNet++ encoder (ablation) | PointNet++ · PCSOD test | IoU: 0.608 -> 0.624 | Improved |
| add PPB 1 (global semantics) on top of SPB (ablation) | PointNet++ +SPB · PCSOD test | IoU: 0.624 -> 0.639 | Improved |
| add PPB 2 (multi-scale) on top of SPB and PPB 1 (ablation) | PointNet++ +SPB +PPB 1 · PCSOD test | IoU: 0.639 -> 0.656 | Improved |
| add PPB 2 on top of SPB and PPB 1 (ablation) | PointNet++ +SPB +PPB 1 · PCSOD test | MAE: 0.073 -> 0.069 | Improved |
| use mean-max reduction instead of mean reduction (ablation) | PPB · PCSOD test | F-measure: 0.764 -> 0.769 | Improved |
| use mean-max reduction instead of max reduction (ablation) | PPB · PCSOD test | F-measure: 0.765 -> 0.769 | Improved |
| use mean-max reduction instead of attentive reduction (ablation) | PPB · PCSOD test | MAE: 0.074 -> 0.069 | Improved |
| use mean-max reduction instead of attentive reduction (ablation) | PPB · PCSOD test | IoU: 0.658 -> 0.656 | Hurt |
| replace PointNet++ with proposed model with PPB and SPB | PointNet++ · PCSOD test | F-measure: 0.738 -> 0.769 | Improved |
| compare RandLA to PointNet++ for SOD | RandLA · PCSOD test | MAE: 0.077 -> 0.127 | 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.
- ShellNet — ShellNet: Efficient Point Cloud Convolutional Neural Networks Using Concentric Shells Statistics (ICCV 2019) · better on 4 of 4 shared results
- PointCNN — PointCNN: Convolution On X-Transformed Points (NeurIPS 2018) · better on 4 of 4 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
- MSV-PCT: Multi-Sparse-View Enhanced Transformer Framework for Salient Object Detection in Point Clouds (AAAI 2025 · similarity 0.91)
- Salient Objects in Clutter: Bringing Salient Object Detection to the Foreground (ECCV 2018 · similarity 0.91)
- Breaking Barriers, Localizing Saliency: A Large-Scale Benchmark and Baseline for Condition-Constrained Salient Object Detection (TPAMI 2025 · similarity 0.91)
- Geometry-Aware 3D Salient Object Detection Network (AAAI 2025 · similarity 0.90)
- Salient Objects in Clutter (TPAMI 2022 · similarity 0.90)
- WeakPCSOD: Overcoming the Bias of Box Annotations for Weakly Supervised Point Cloud Salient Object Detection (AAAI 2024 · similarity 0.88)
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.
