Research map / Label-efficient and robust learning / Domain adaptation and generalization
Spatial Attention Pyramid Network for Unsupervised Domain Adaptation
Congcong Li, Dawei Du, Libo Zhang, Longyin Wen, Tiejian Luo, Yanjun Wu, Pengfei Zhu
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 task-specific guided map (ablation) | SAPNet · PASCAL VOC->Clipart | mAP: 37.1 -> 42.2 | Improved |
| add task-specific guided map (ablation) | SAPNet · Sim10k->Cityscapes | AP: 43.8 -> 44.9 | Improved |
| add task-specific guided map (ablation) | SAPNet · Cityscapes->FoggyCityscapes | mAP: 38.3 -> 40.9 | Improved |
| add spatial attention pyramid (N=13) (ablation) | SAPNet · PASCAL VOC->Clipart | mAP: 35.4 -> 42.2 | Improved |
| add spatial attention pyramid (N=13) (ablation) | SAPNet · Sim10k->Cityscapes | AP: 38.3 -> 44.9 | Improved |
| add spatial attention pyramid (N=13) (ablation) | SAPNet · Cityscapes->FoggyCityscapes | mAP: 36.6 -> 40.9 | Improved |
| increase pyramid levels from N=3 to N=13 (ablation) | SAPNet · PASCAL VOC->Clipart | mAP: 39.6 -> 42.2 | Improved |
| increase pyramid levels from N=7 to N=13 (ablation) | SAPNet · Sim10k->Cityscapes | AP: 45.9 -> 44.9 | Hurt |
| use average pooling instead of max pooling (ablation) | SAPNet · Cityscapes->FoggyCityscapes | mAP: 34.9 -> 40.9 | Improved |
| use average pooling instead of max pooling (ablation) | SAPNet · PASCAL VOC->Clipart | mAP: 37.5 -> 42.2 | Improved |
| add channel-wise attention weighting (ablation) | SAPNet · PASCAL VOC->Clipart | mAP: 37.7 -> 42.2 | Improved |
| add channel-wise attention weighting (ablation) | SAPNet · Sim10k->Cityscapes | AP: 45.6 -> 44.9 | Hurt |
Closest papers by meaning
- Unsupervised Domain Adaptation of Object Detectors: A Survey (TPAMI 2023 · similarity 0.90)
- Deep Adversarial Attention Alignment for Unsupervised Domain Adaptation: the Benefit of Target Expectation Maximization (ECCV 2018 · similarity 0.89)
- DAST: Unsupervised Domain Adaptation in Semantic Segmentation Based on Discriminator Attention and Self-Training (AAAI 2021 · similarity 0.89)
- CDAC: Cross-domain Attention Consistency in Transformer for Domain Adaptive Semantic Segmentation (ICCV 2023 · similarity 0.88)
- Self-Ensembling Attention Networks: Addressing Domain Shift for Semantic Segmentation (AAAI 2019 · similarity 0.88)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · 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.
