Research map / Label-efficient and robust learning / Domain adaptation and generalization
Progressive Adversarial Networks for Fine-Grained Domain Adaptation
Check 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 |
|---|---|---|---|
| apply progressive adversarial adaptation with progressively mixed coarse-to-fine labels | ResNet-50 domain adaptation · CUB-Paintings | average accuracy: 50.28% to 59.16% | Helped |
| apply PAN instead of the strongest reported baseline | ResNet-50 domain adaptation · CompCars | average accuracy: 29.13 to 31.31 | Helped |
| apply PAN instead of the strongest reported baseline | ResNet-50 domain adaptation · Birds-31 | average accuracy: 82.18 to 83.34 | Helped |
| add progressive granularity learning (PGL) to PAL (ablation) | PAN on CUB-Paintings · CUB-Paintings | average accuracy: 54.44 to 59.16 | Helped |
| add progressive adversarial learning (PAL) to PGL (ablation) | PAN on CUB-Paintings · CUB-Paintings | average accuracy: 53.87 to 59.16 | Helped |
| add residual feature concatenation to PAL (ablation) | PAN on CUB-Paintings · CUB-Paintings | average accuracy: 53.89 to 59.16 | Helped |
| share weights between the fine- and coarse-grained feature extractors (ablation) | PAN · CUB-Paintings | average accuracy: 59.16% to 51.48% | Hurt |
| use coarse labels at genus, family, and order levels together rather than one level alone (ablation) | PAN · CUB-Paintings | average accuracy: 56.85, 56.76, and 57.88 to 59.16 | Helped |
| replace the proposed curriculum schedule with a linear schedule (ablation) | PAN · CUB-Paintings | average accuracy: 59.16 to 54.67 | Hurt |
| replace the proposed curriculum schedule with a step schedule (ablation) | PAN · CUB-Paintings | average accuracy: 59.16 to 54.88 | Hurt |
| replace the proposed curriculum schedule with an exponential schedule (ablation) | PAN · CUB-Paintings | average accuracy: 59.16 to 55.19 | Hurt |
Closest papers by meaning
- Adversarial Discriminative Domain Adaptation (CVPR 2017 · similarity 0.89)
- Reusing the Task-Specific Classifier as a Discriminator: Discriminator-Free Adversarial Domain Adaptation (CVPR 2022 · similarity 0.88)
- f-Domain Adversarial Learning: Theory and Algorithms (ICML 2021 · similarity 0.88)
- Multi-Adversarial Domain Adaptation (AAAI 2018 · similarity 0.87)
- Boosting Domain Adaptation by Discovering Latent Domains (CVPR 2018 · similarity 0.87)
- GCAN: Graph Convolutional Adversarial Network for Unsupervised Domain Adaptation (CVPR 2019 · similarity 0.87)
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.