Research map / Label-efficient and robust learning / Domain adaptation and generalization
Generalized Domain Adaptation
Yu Mitsuzumi, Go Irie, Daiki Ikami, Takashi Shibata
Check 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 |
|---|---|---|---|
| Use class-destructive self-supervised domain estimation followed by domain-adversarial classifier learning | GDA1 · Digits (Avg.) | HOS: 80.07 vs 37.46 | Improved |
| Use class-destructive self-supervised domain estimation followed by domain-adversarial classifier learning | GDA1 · Office-31 (Avg.) | HOS: 81.02 vs 64.31 | Improved |
| Use class-destructive self-supervised domain estimation followed by domain-adversarial classifier learning | GDA2 · Digits (Avg.) | HOS: 76.70 vs 51.74 | Improved |
| Use class-destructive self-supervised domain estimation followed by domain-adversarial classifier learning | GDA2 · Office-31 (Avg.) | HOS: 79.49 vs 71.19 | Improved |
| Apply the general GDA method rather than an OSDA-specific baseline | Office-31 OSDA · Office-31 (Avg.) | HOS: 85.9 vs 85.3 | Hurt |
| Apply the general GDA method rather than an OSDA-specific baseline | Office-Home OSDA · Office-Home (Avg.) | HOS: 66.2 vs 65.8 | Hurt |
| Apply the general GDA method rather than the specialized MS-OSDA baseline | Office-31 MS-OSDA · Office-31 (Avg.) | OS*: 93.1 vs 88.3 | Hurt |
| Apply the general GDA method rather than the specialized BTDA baseline | Office-31 BTDA · Office-31 (Avg.) | accuracy: 80.2 vs 78.8 | Hurt |
| Increase the number of grid partitions in the class-destructive image transformation (ablation) | GDA domain-label estimation · Digits: sv(0-4), sy(4-7) | accuracy | Hurt |
| Use more clusters than the true number of domains (ablation) | GDA domain-label estimation · Digits: sv(0,1), sy(2,3), mt(4,5), mm(6,7) | HOS | No clear effect |
Closest papers by meaning
- Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift (NeurIPS 2020 · similarity 0.91)
- Divergence-Agnostic Unsupervised Domain Adaptation by Adversarial Attacks (TPAMI 2021 · similarity 0.90)
- GCAN: Graph Convolutional Adversarial Network for Unsupervised Domain Adaptation (CVPR 2019 · similarity 0.90)
- Unsupervised Multi-Class Domain Adaptation: Theory, Algorithms, and Practice (TPAMI 2020 · similarity 0.89)
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.89)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · 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.
