Research map / Label-efficient and robust learning / Domain adaptation and generalization
Deep Asymmetric Transfer Network for Unbalanced Domain Adaptation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 unsupervised transfer (distribution matching on unlabeled data) (ablation) | DATN · NUS-WIDE (SIFT) | accuracy: 0.358 -> 0.375 | Helped |
| add unsupervised transfer (ablation) | DATN · NUS-WIDE (VGG16) | accuracy: 0.825 -> 0.861 | Helped |
| add unsupervised transfer (ablation) | DATN · Amazon EN->FR | accuracy: 0.724 -> 0.737 | Helped |
| add unsupervised transfer (ablation) | DATN · Amazon EN->JP | accuracy: 0.734 -> 0.755 | Helped |
| add asymmetric supervised transfer to unsupervised-only model (ablation) | DATN · NUS-WIDE (SIFT) | accuracy: 0.285 -> 0.375 | Helped |
| add asymmetric supervised transfer to unsupervised-only model (ablation) | DATN · Amazon EN->FR | accuracy: 0.638 -> 0.737 | Helped |
| asymmetric transfer (DATN sup) instead of symmetric transfer (WSDTN) (ablation) | DATN · Amazon EN->FR | accuracy: 0.675 -> 0.724 | Helped |
| asymmetric adaptive transfer instead of fully transferring source knowledge (HHTL) (ablation) | DATN · Amazon EN->GE | accuracy: 0.673 -> 0.718 | Helped |
| add unlabeled source data when set is small (ablation) | DATN · NUS-WIDE VGG16 / Amazon EN-FR | accuracy | Mixed |
| increase number of co-occurred domain pairs (ablation) | DATN · all transfer methods | accuracy | Helped |
Closest papers by meaning
- Unsupervised Domain Adaptation with Residual Transfer Networks (NeurIPS 2016 · similarity 0.91)
- Residual Parameter Transfer for Deep Domain Adaptation (CVPR 2018 · similarity 0.89)
- Learning to Transfer Examples for Partial Domain Adaptation (CVPR 2019 · similarity 0.89)
- Domain-Constraint Transfer Coding for Imbalanced Unsupervised Domain Adaptation (AAAI 2016 · similarity 0.89)
- Domain-Symmetric Networks for Adversarial Domain Adaptation (CVPR 2019 · similarity 0.88)
- Learning Transferrable Representations for Unsupervised Domain Adaptation (NeurIPS 2016 · similarity 0.88)
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.