Research map / Label-efficient and robust learning / Domain adaptation and generalization
Prototype-Guided Continual Adaptation for Class-Incremental Unsupervised Domain Adaptation
Hongbin Lin, Yifan Zhang, Zhen Qiu, Shuaicheng Niu, Chuang Gan, Yanxia Liu, Mingkui Tan
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 L_ce on pseudo-labeled target data (ablation) | backbone · Office-31-CI | Final accuracy: 77.5 -> 79.4 | Improved |
| add prototype knowledge replay loss L_dis (ablation) | ProCA (L_ce) · Office-31-CI | Final accuracy: 79.4 -> 81.2 | Improved |
| add prototype knowledge replay loss L_dis (ablation) | ProCA (L_ce) · Office-31-CI | Final S-1 accuracy: 86.8 -> 89.0 | Improved |
| add prototype contrastive alignment L_con (ablation) | ProCA (L_ce) · Office-31-CI | Final accuracy: 79.4 -> 80.8 | Improved |
| add prototype contrastive alignment L_con (ablation) | ProCA (L_ce) · Office-31-CI | Final S-1 accuracy: 86.8 -> 88.0 | Improved |
| add L_con on top of L_ce and L_dis (ablation) | ProCA · Office-31-CI | Final accuracy: 81.2 -> 82.1 | Improved |
| add L_dis on top of L_ce and L_con (ablation) | ProCA · Office-31-CI | Final accuracy: 80.8 -> 82.1 | Improved |
| add L_con on top of L_ce and L_dis (ablation) | ProCA · Office-31-CI | Final S-1 accuracy: 89.0 -> 90.8 | Improved |
| add label prototype identification (ablation) | PADA · Office-31-CI | Final accuracy: 57.4 -> 73.0 | Improved |
| add label prototype identification (ablation) | PADA · Office-31-CI | Final S-1 accuracy: 48.3 -> 80.4 | Improved |
| add label prototype identification (ablation) | BA3US · Office-31-CI | Final accuracy: 78.3 -> 80.4 | Improved |
| add label prototype identification (ablation) | BA3US · Office-31-CI | Final S-1 accuracy: 88.7 -> 89.8 | Improved |
Closest papers by meaning
- A Prototype-Oriented Framework for Unsupervised Domain Adaptation (NeurIPS 2021 · similarity 0.91)
- Class-Incremental Domain Adaptation (ECCV 2020 · similarity 0.91)
- Prior Knowledge Guided Unsupervised Domain Adaptation (ECCV 2022 · similarity 0.91)
- Partial Label Unsupervised Domain Adaptation with Class-Prototype Alignment (ICLR 2023 · similarity 0.90)
- Progressive Feature Alignment for Unsupervised Domain Adaptation (CVPR 2019 · similarity 0.89)
- Subsidiary Prototype Alignment for Universal Domain Adaptation (NeurIPS 2022 · 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.
