Research map / Label-efficient and robust learning / Domain adaptation and generalization
Representation Learning with Multiple Lipschitz-Constrained Alignments on Partially-Labeled Cross-Domain Data
Songlei Jian, Liang Hu, Longbing Cao, Kai Lu
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 complementary representation with self alignment loss LSA (ablation) | MULANCom · VisDA semi-supervised DA, 10% labeled | accuracy (%): 77.44±0.2 -> 78.11±0.1 | Improved |
| add self alignment LSA (ablation) | MULANCom · VisDA semi-supervised DA, 20% labeled | accuracy (%): 78.48±0.2 -> 79.15±0.2 | Improved |
| add self alignment LSA (ablation) | MULANCom · VisDA semi-supervised DA, 50% labeled | accuracy (%): 79.76±0.1 -> 80.64±0.1 | Improved |
| use common representation (class + distribution alignment) vs TADT baseline | MULANCom · VisDA semi-supervised DA, 10% labeled | accuracy (%): 69.68±0.4 -> 77.44±0.2 | Improved |
| increase labeled source data from 10% to 50% | Source-only · VisDA semi-supervised DA | accuracy (%): 45.01±1.0 -> 38.82±0.5 | Hurt |
| increase labeled data from 10% to 50% | MULAN · VisDA semi-supervised DA | accuracy (%): 78.11±0.1 -> 80.64±0.1 | Improved |
| use complementary representation with target classifier instead of common representation with source classifier (ablation) | MULAN · VisDA few-shot DA, 5-shot | accuracy (%): 56.02 -> 57.97 | Improved |
| use complementary representation instead of common representation (ablation) | MULAN · VisDA few-shot DA, 10-shot | accuracy (%): 58.17 -> 62.29 | Improved |
| use complementary representation instead of common representation (ablation) | MULAN · VisDA few-shot DA, 20-shot | accuracy (%): 64.40 -> 66.73 | Improved |
| MULAN complementary representation vs Target-only classifier | MULAN · VisDA few-shot DA, 5-shot | accuracy (%): 41.17 -> 57.97 | Improved |
| reduce shots from 20 to 5 | MULAN(Comp) · VisDA few-shot DA | accuracy (%): 66.73 -> 57.97 | Hurt |
| reduce shots from 20 to 5 | DANN · VisDA few-shot DA | accuracy (%): 38.21 -> 35.27 | Hurt |
Closest papers by meaning
- Maximum Structural Generation Discrepancy for Unsupervised Domain Adaptation (TPAMI 2022 · similarity 0.88)
- ECACL: A Holistic Framework for Semi-Supervised Domain Adaptation (ICCV 2021 · similarity 0.88)
- OSAN: A One-Stage Alignment Network To Unify Multimodal Alignment and Unsupervised Domain Adaptation (CVPR 2023 · similarity 0.87)
- Forget More to Learn More: Domain-specific Feature Unlearning for Semi-supervised and Unsupervised Domain Adaptation (ECCV 2024 · similarity 0.87)
- A General Representation-Based Approach to Multi-Source Domain Adaptation (ICML 2025 · similarity 0.86)
- Tackling Dimensional Collapse toward Comprehensive Universal Domain Adaptation (ICML 2025 · similarity 0.86)
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.
