Research map / Label-efficient and robust learning / Domain adaptation and generalization
Weak Adaptation Learning: Addressing Cross-Domain Data Insufficiency With Weak Annotator
Shichao Xu, Lixu Wang, Yixuan Wang, Qi 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 |
|---|---|---|---|
| train with weak annotator labels plus WAL stages instead of target data only | B_t · CIFAR-10 | accuracy: 30.46 -> 61.71 | Improved |
| apply WAL instead of fine-tuning all parameters | B_f2 · M->S digits | accuracy: 77.92 -> 80.00 | Improved |
| apply WAL instead of target-only training | B_t · VisDA-C | accuracy: 32.86 -> 40.83 | Improved |
| apply WAL instead of best SSDA method FAN | FAN · VisDA-C | accuracy: 32.99 -> 40.83 | Improved |
| fine-tune only last three layers on weak-label pretrained model | B_f1 · S->M digits | accuracy: 94.79 -> 80.41 | Hurt |
| fine-tune all layers instead of last three (ablation) | B_f1 · VisDA-C | accuracy: 27.67 -> 35.03 | Improved |
| use a much weaker annotator (23.79 accuracy) (ablation) | WAL · CIFAR-10 | accuracy: 61.71 -> 42.29 | Hurt |
| increase number of labeled target samples (ablation) | WAL · CIFAR-10 | accuracy | Improved |
| increase accuracy of weak annotator (ablation) | WAL · CIFAR-10 | accuracy | Improved |
| add unlabeled target samples via SSDA methods (FAN, MME, ENT) | SSDA baselines · VisDA-C | accuracy: 32.99 | No clear effect |
Closest papers by meaning
- A Review of Domain Adaptation without Target Labels (TPAMI 2019 · similarity 0.89)
- Transferable Curriculum for Weakly-Supervised Domain Adaptation (AAAI 2019 · similarity 0.88)
- Training Subset Selection for Weak Supervision (NeurIPS 2022 · similarity 0.88)
- Weakly Supervised Open-Set Domain Adaptation by Dual-Domain Collaboration (CVPR 2019 · similarity 0.88)
- Gradient-Based Sample Selection for Black-Box Universal Domain Adaptation (AAAI 2025 · similarity 0.87)
- Universal Semi-Supervised Domain Adaptation by Mitigating Common-Class Bias (CVPR 2024 · similarity 0.86)
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.
