Research map / Label-efficient and robust learning / Domain adaptation and generalization
Hard-Aware Instance Adaptive Self-Training for Unsupervised Cross-Domain Semantic Segmentation
Chuang Zhu, Kebin Liu, Wenqi Tang, Ke Mei, Jiaqi Zou, Tiejun Huang
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 instance adaptive selector (IAS) (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 43.8 -> 49.5 | Improved |
| add confident region KLD minimization (Rc) (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 49.5 -> 50.5 | Improved |
| add ignored region entropy minimization (Ri) (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 50.5 -> 51.5 | Improved |
| add hard-aware pseudo-label augmentation (HPLA) (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 51.5 -> 55.7 | Improved |
| add ignored region consistency constraint (Rcst) (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 55.7 -> 56.3 | Improved |
| add EMA in IAS (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 55.2 -> 56.3 | Improved |
| add warm-up model before self-training (ablation) | HIAST · GTA5 -> Cityscapes | mIoU: 35.6 -> 43.8 | Improved |
| replace AdaptSeg with SePiCo as warm-up model | HIAST · GTA5 -> Cityscapes | mIoU: 56.3 -> 64.1 | Improved |
| apply HIAST to DAFormer | DAFormer · GTA5 -> Cityscapes | mIoU: 68.3 -> 69.5 | Improved |
| apply HIAST to HRDA | HRDA · GTA5 -> Cityscapes | mIoU: 73.8 -> 74.5 | Improved |
| apply HIAST to AdaptSeg (one round) | AdaptSeg · GTA5 -> Cityscapes | mIoU: 42.4 -> 53.6 | Improved |
| apply HIAST to CPSL (one round) | CPSL · GTA5 -> Cityscapes | mIoU: 60.8 -> 63.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DAFormer — DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation (CVPR 2022) · better on 1 of 1 shared result
- HRDA — HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation (ECCV 2022) · better on 1 of 1 shared result
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
- Instance Adaptive Self-Training for Unsupervised Domain Adaptation (ECCV 2020 · similarity 0.96)
- Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training (ECCV 2018 · similarity 0.93)
- Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentation (IJCV 2021 · similarity 0.92)
- Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (CVPR 2021 · similarity 0.92)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · similarity 0.90)
- Balanced Learning for Domain Adaptive Semantic Segmentation (ICML 2025 · similarity 0.90)
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.
