Research map / Detection and segmentation / Open-set and few-shot detection
Amodal Segmentation Based on Visible Region Segmentation and Shape Prior
Yuting Xiao, Yanyu Xu, Ziming Zhong, Weixin Luo, Jiawei Li, Shenghua Gao
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 |
|---|---|---|---|
| use visible-region features and category shape priors to infer amodal masks | Our method · D2SA, KINS, COCOA cls | Amodal AP: 64.85 -> 70.27; 30.71 -> 32.08; 34.72 -> 35.41 | Improved |
| add visible attention (ablation) | Our method · D2SA; KINS | Amodal AP: 66.67 -> 69.98; 31.61 -> 31.94 | Improved |
| add the reclassification regularizer (ablation) | Our method · D2SA; KINS | Amodal AP: 69.02 -> 69.98; 31.65 -> 31.94 | Improved |
| add shape-prior refinement (ablation) | Our method · D2SA; KINS | Amodal AP: 68.84 -> 69.98; 31.70 -> 31.94 | Improved |
| add shape-prior post-processing (ablation) | Our method · D2SA; KINS | Amodal AP: 68.15 -> 69.98; 31.87 -> 31.94 | Improved |
| add feature matching to both mask heads (ablation) | Our method · D2SA; KINS | Amodal AP: 69.98 -> 70.27; 31.94 -> 32.08 | Improved |
| use amodal attention instead of visible attention for amodal refinement (ablation) | attention variants · D2SA; KINS | Amodal AP: 64.55 vs 66.98; 31.18 vs 31.32 | Hurt |
| use refined rather than coarse visible attention for amodal refinement (ablation) | attention variants · D2SA; KINS | Amodal AP: 67.11 -> 67.33; 31.57 -> 31.69 | Improved |
Closest papers by meaning
- Using Diffusion Priors for Video Amodal Segmentation (CVPR 2025 · similarity 0.90)
- Amodal Ground Truth and Completion in the Wild (CVPR 2024 · similarity 0.90)
- BLADE: Box-Level Supervised Amodal Segmentation through Directed Expansion (AAAI 2024 · similarity 0.90)
- Segment Anything, Even Occluded (CVPR 2025 · similarity 0.89)
- Towards Efficient Foundation Model for Zero-shot Amodal Segmentation (CVPR 2025 · similarity 0.88)
- Amodal Scene Analysis via Holistic Occlusion Relation Inference and Generative Mask Completion (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.
