Research map / Label-efficient and robust learning / Domain adaptation and generalization
Multi-Source Open-Set Deep Adversarial Domain Adaptation
Sayan Rakshit, Dipesh Tamboli, Pragati Shuddhodhan Meshram, Biplab Banerjee, Gemma Roig, Subhasis Chaudhuri
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 margin loss L_margin (ablation) | MOSDANet · Office-31 A,D-W and Office-Home ACR-P | OS (%): at least by 4% | Improved |
| add margin loss L_margin (ablation) | MOSDANet · Office-31 A,D-W | known vs unknown probability difference: ≈0.3 -> ≈0.55 | Improved |
| raise margin tau from 0.2 to 0.6 (ablation) | MOSDANet · Office-Home ACR-P | OS (%): 66.4 -> 80.0 | Improved |
| raise margin tau from 0.2 to 0.6 (ablation) | MOSDANet · Office-31 AD-W | OS (%): 95.8 -> 98.2 | Improved |
| add explicit source-domain alignment term (ablation) | MOSDANet · Office-31 A,D-W | OS and OS* (%): more than 10% | Improved |
| use ensemble decision fusion instead of best member classifier (ablation) | MOSDANet · Office-31 (avg) | OS (%): 84.9 -> 91.9 | Improved |
| use ensemble decision fusion instead of best member classifier (ablation) | MOSDANet · Office-Caltech (avg) | OS (%): 92.9 -> 95.8 | Improved |
| use ensemble decision fusion instead of best member classifier (ablation) | MOSDANet · Office-Home (avg) | OS (%): 70.5 -> 72.1 | Improved |
| multi-source MOSDANet vs source-combine IOSDA-BP | IOSDA-BP · Office-31 (avg) | OS (%): 83.2 -> 91.9 | Improved |
| increase openness from O=0.03 to O=0.35 (ablation) | MOSDANet · Office-31 AD-W | OS (%): 89.4 -> 98.2 | Improved |
| closed-set DA (DANN) alignment before OSVM vs OSVM alone | OSVM (source combine) · Office-31 (avg) | OS (%): 56.3 -> 70.9 | Improved |
Closest papers by meaning
- Unknown-Aware Domain Adversarial Learning for Open-Set Domain Adaptation (NeurIPS 2022 · similarity 0.91)
- Bridging Theory and Practice in Source-Free Domain Adaptation via Adversarial Proxy Perturbation (ECCV 2026 · similarity 0.90)
- Adaptive Adversarial Network for Source-Free Domain Adaptation (ICCV 2021 · similarity 0.90)
- Adversarial Multiple Source Domain Adaptation (NeurIPS 2018 · similarity 0.90)
- Open Set Domain Adaptation by Backpropagation (ECCV 2018 · similarity 0.90)
- Multi-Adversarial Domain Adaptation (AAAI 2018 · similarity 0.90)
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.
