Research map / Detection and segmentation / Open-set and few-shot detection
Coarse-to-Fine Amodal Segmentation with Shape Prior
Jianxiong Gao, Xuelin Qian, Yikai Wang, Tianjun Xiao, Tong He, Zheng Zhang, Yanwei Fu
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 convolutional refinement module (ablation) | C2F-Seg · COCOA | mIoU_occ: 24.25 -> 27.71 | Improved |
| add convolutional refinement module (ablation) | C2F-Seg · KINS | mIoU_occ: 52.57 -> 53.60 | Improved |
| predict both visible and amodal masks in the refine module (two branches) (ablation) | C2F-Seg · COCOA | mIoU_occ: 26.12 -> 27.71 | Improved |
| add attention computed from coarse mask in refine module (ablation) | C2F-Seg · COCOA | mIoU_occ: 26.85 -> 27.71 | Improved |
| add spatial-temporal transformer block (ablation) | C2F-Seg video · MOViD-A | mIoU_occ: 26.48 -> 36.13 | Improved |
| add temporal roll by T/2 frames after each block (ablation) | C2F-Seg video · MOViD-A | mIoU_occ: 32.35 -> 36.13 | Improved |
| increase iterative decoding steps K (ablation) | C2F-Seg · COCOA and MOViD-A | mIoU: COCOA mIoU_occ 27.70 (K=1) vs 27.71 (K=3) | No clear effect |
| feed ground-truth visible masks instead of predicted ones (ablation) | C2F-Seg · COCOA | mIoU_occ: 27.71 -> 36.55 | Improved |
| use VQ-GAN latent for the image instead of pretrained ResNet features (ablation) | C2F-Seg · amodal segmentation | performance | Hurt |
| apply C2F-Seg versus VRSP on COCOA AP75 | C2F-Seg · COCOA | AP75: 38.7 -> 38.5 | No clear effect |
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 0.89)
- Towards Efficient Foundation Model for Zero-shot Amodal Segmentation (CVPR 2025 · similarity 0.88)
- BLADE: Box-Level Supervised Amodal Segmentation through Directed Expansion (AAAI 2024 · similarity 0.88)
- Segment Anything, Even Occluded (CVPR 2025 · similarity 0.88)
- pix2gestalt: Amodal Segmentation by Synthesizing Wholes (CVPR 2024 · 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.
