Research map / Label-efficient and robust learning / Domain adaptation and generalization
MADAv2: Advanced Multi-Anchor Based Active Domain Adaptation Segmentation
Munan Ning, Donghuan Lu, Yujia Xie, Dongdong Chen, Dong Wei, Yefeng Zheng, Yonghong Tian, Shuicheng Yan, Li Yuan
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 active target samples (5%) with cross-entropy (ablation) | MADAv2 baseline · GTA5 to Cityscapes | mIoU: 42.5 -> 65.1 | Improved |
| add consistency loss and self-training (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 65.1 -> 68.5 | Improved |
| add OHEM loss (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 68.5 -> 69.5 | Improved |
| add ACM and ACP augmentation (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 69.5 -> 70.8 | Improved |
| add soft-anchor alignment loss (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 70.8 -> 71.4 | Improved |
| add active samples (ablation) | MADAv2 baseline · SYNTHIA to Cityscapes | mIoU: 42.9 -> 65.2 | Improved |
| add soft-anchor alignment loss (ablation) | MADAv2 · SYNTHIA to Cityscapes | mIoU: 71.1 -> 71.8 | Improved |
| select samples with Dual_Domain_Distance instead of single-centroid prototype distance (ablation) | MADAv2 selection · GTA5 to Cityscapes | mIoU: 63.1 -> 65.1 | Improved |
| select samples with Dual_Domain_Distance instead of random (ablation) | MADAv2 selection · GTA5 to Cityscapes | mIoU: 61.1 -> 65.1 | Improved |
| use MADAv2 sample selection instead of MADA selection with MADAv2 semi-supervision (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 70.5 -> 71.4 | Improved |
| use MADAv2 semi-supervision instead of MADA with MADA selection (ablation) | MADA selection · GTA5 to Cityscapes | mIoU: 66.9 -> 70.5 | Improved |
| use only target anchor distance and only target data (drop source information) (ablation) | MADAv2 · GTA5 to Cityscapes | mIoU: 71.4 -> 67.3 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- HRDA — HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation (ECCV 2022) · better on 2 of 2 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
- Multi-Anchor Active Domain Adaptation for Semantic Segmentation (ICCV 2021 · similarity 0.94)
- D2ADA: Dynamic Density-Aware Active Domain Adaptation for Semantic Segmentation (ECCV 2022 · similarity 0.90)
- Learning to Adapt via Latent Domains for Adaptive Semantic Segmentation (NeurIPS 2021 · similarity 0.88)
- Category-Aware Active Domain Adaptation (ICML 2024 · similarity 0.88)
- Revisiting the Domain Shift and Sample Uncertainty in Multi-source Active Domain Transfer (CVPR 2024 · similarity 0.88)
- Density Matters: Improved Core-Set for Active Domain Adaptive Segmentation (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.
