Research map / Label-efficient and robust learning / Domain adaptation and generalization
Contextual-Relation Consistent Domain Adaptation for Semantic Segmentation
Jiaxing Huang, Shijian Lu, Dayan Guan, Xiaobing Zhang
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 pixel-scale adaptation (AEMM) (ablation) | ResNet-101 baseline · GTA5->Cityscapes (ResNet-101) | mIoU: 36.6 -> 43.8 | Improved |
| use AEMM instead of MinEnt at pixel scale (ablation) | pixel-scale adaptation · GTA5->Cityscapes (ResNet-101) | mIoU: 42.4 -> 43.8 | Improved |
| use AEMM instead of MinEnt at local scale (ablation) | CrCDA* · GTA5->Cityscapes (ResNet-101) | mIoU: 42.1 -> 43.7 | Improved |
| add local-scale adaptation to pixel+global (ablation) | CrCDA · GTA5->Cityscapes (ResNet-101) | mIoU: 46.0 -> 48.6 | Improved |
| add global-scale adaptation to pixel+local (ablation) | CrCDA · GTA5->Cityscapes (ResNet-101) | mIoU: 45.6 -> 48.6 | Improved |
| add pixel-scale adaptation to local+global (ablation) | CrCDA · GTA5->Cityscapes (ResNet-101) | mIoU: 46.1 -> 48.6 | Improved |
| add local-scale adaptation to AdaptSeg (ablation) | AdaptSeg · GTA5->Cityscapes (ResNet-101) | mIoU: 41.4 -> 44.8 | Improved |
| add local-scale adaptation to PatAlign (ablation) | PatAlign · GTA5->Cityscapes (ResNet-101) | mIoU: 41.3 -> 44.7 | Improved |
| add local-scale adaptation to AdvEnt (ablation) | AdvEnt · GTA5->Cityscapes (ResNet-101) | mIoU: 43.8 -> 45.2 | Improved |
| add local-scale adaptation to our global alignment (ablation) | GlobalAlign · GTA5->Cityscapes (ResNet-101) | mIoU: 44.3 -> 46.1 | Improved |
| combine AdaptSeg and PatAlign global methods (ablation) | AdaptSeg · GTA5->Cityscapes (ResNet-101) | mIoU: 41.4 -> 43.2 | Improved |
| use CrCDA instead of PatAlign on SYNTHIA (ResNet-101) | PatAlign · SYNTHIA->Cityscapes (ResNet-101) | mIoU*: 46.5 -> 50.0 | Improved |
Closest papers by meaning
- Taking a Closer Look at Domain Shift: Category-Level Adversaries for Semantics Consistent Domain Adaptation (CVPR 2019 · similarity 0.92)
- Category-Level Adversarial Adaptation for Semantic Segmentation using Purified Features (TPAMI 2021 · similarity 0.91)
- An Adversarial Perturbation Oriented Domain Adaptation Approach for Semantic Segmentation (AAAI 2020 · similarity 0.91)
- I2F: A Unified Image-to-Feature Approach for Domain Adaptive Semantic Segmentation (TPAMI 2022 · similarity 0.90)
- Consistency Regularization with High-dimensional Non-adversarial Source-guided Perturbation for Unsupervised Domain Adaptation in Segmentation (AAAI 2021 · similarity 0.90)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · similarity 0.89)
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.
