Research map / Label-efficient and robust learning / Domain adaptation and generalization
Overcoming Shortcut Learning in a Target Domain by Generalizing Basic Visual Factors from a Source Domain
Piyapat Saranrittichai, Chaithanya Kumar Mummadi, Claudia Blaiotta, Mauricio Munoz, Volker Fischer
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 source domain with IL constraint | Factor-0 · DiagVib-Animal | HM accuracy: 1.5 -> 41.3 | Improved |
| add source domain with IL constraint | Factor-0 · Color-Fruit | HM accuracy: 5.5 -> 57.0 | Improved |
| add source domain with IL constraint | Factor-0 · AO-CLEVR | HM accuracy: 4.3 -> 32.1 | Improved |
| add source domain with IL constraint | Factor-0 · Color-Fashion | HM accuracy: 3.6 -> 22.3 | Improved |
| use factor instead of global representation with source (ablation) | FactorSRC · Color-Fruit | HM accuracy: 2.2 -> 42.4 | Improved |
| add isolated latent constraint to FactorSRC (ablation) | FactorSRC · DiagVib-Animal | HM accuracy: 13.0 -> 41.3 | Improved |
| add cross-factor independence constraint to FactorSRC (ablation) | FactorSRC · Color-Fruit | HM accuracy: 42.4 -> 19.6 | Hurt |
| add IL constraint (ablation) | FactorSRC · DiagVib-Animal cross-prediction z_a to object | cross-prediction accuracy (lower is better): 77 -> 46 | Improved |
| add regularization when learning association matrix (ablation) | FactorSRC-IL-LA · DiagVib-Animal (semi-correlated) | HM accuracy: 7.1 -> 34.7 | Improved |
| make source factors uncorrelated (ablation) | FactorSRC-IL · DiagVib-Animal | HM accuracy: 2.5 -> 33.5 | Improved |
| use uncorrelated Caltech source instead of correlated target data (ablation) | FactorSRC-IL · DiagVib-Animal | HM accuracy: 1.7 -> 33.5 | Improved |
| use Caltech shapes instead of MNIST shapes in source (ablation) | FactorSRC-IL · Color-Fruit | HM accuracy: 46.9 -> 56.0 | Improved |
Closest papers by meaning
- On the Foundations of Shortcut Learning (ICLR 2024 · similarity 0.87)
- When and How Does CLIP Enable Domain and Compositional Generalization? (ICML 2025 · similarity 0.87)
- Improved Training Technique for Shortcut Models (NeurIPS 2025 · similarity 0.87)
- Do ImageNet-trained Models Learn Shortcuts? The Impact of Frequency Shortcuts on Generalization (CVPR 2025 · similarity 0.86)
- When does compositional structure yield compositional generalization? A kernel theory. (ICLR 2025 · similarity 0.86)
- START: A Generalized State Space Model with Saliency-Driven Token-Aware Transformation (NeurIPS 2024 · similarity 0.85)
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.
