Research map / Label-efficient and robust learning / Domain adaptation and generalization
Unsupervised Domain Adaptation Based on Source-Guided Discrepancy
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 |
|---|---|---|---|
| Use source-guided discrepancy instead of d_H to rank the toy sources | Source selection for the toy target domain · Toy source selection (S1 vs. S2) | target loss of the selected source classifier: 0.49 -> 0.0 | Helped |
| Rank sources using d_H, which does not use source labels | Source selection for the toy target domain · Toy source selection (S1 vs. S2) | target loss of the selected source classifier: 0.0 -> 0.49 | Hurt |
| Use S-disc rather than d_H for source selection among noisy and clean MNIST-M domains | Logistic regression source selection · MNIST-M to MNIST | clean sources ranked in the first five | Helped |
| Use d_H rather than S-disc for source selection among noisy and clean MNIST-M domains | Logistic regression source selection · MNIST-M to MNIST | clean sources ranked in the first five | Hurt |
| Estimate discrepancy with S-disc rather than d_H | Empirical discrepancy estimation · MNIST odd/even classification; S1 and S2 to MNIST | convergence to the true discrepancy | Helped |
| Use S-disc instead of X-disc for discrepancy estimation | Binary classification domain discrepancy estimation · Unsupervised domain adaptation | computational complexity: O((n_T + n_S + d)^8) -> O((n_T + n_S)^3) | Helped |
Closest papers by meaning
- Implicit Task-Driven Probability Discrepancy Measure for Unsupervised Domain Adaptation (NeurIPS 2021 · similarity 0.91)
- Unsupervised Domain Adaptation by Matching Distributions Based on the Maximum Mean Discrepancy via Unilateral Transformations (AAAI 2019 · similarity 0.90)
- Denoised Maximum Classifier Discrepancy for Source-Free Unsupervised Domain Adaptation (AAAI 2022 · similarity 0.89)
- Discrepancy-Based Active Learning for Domain Adaptation (ICLR 2022 · similarity 0.89)
- On $f$-Divergence Principled Domain Adaptation: An Improved Framework (NeurIPS 2024 · similarity 0.88)
- Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · 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.