Research map / Label-efficient and robust learning / Domain adaptation and generalization
Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
Muhammad Ghifary, David Balduzzi, Willem Bastiaan Kleijn, Mengjie Zhang
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 |
|---|---|---|---|
| use label information (SCA instead of unsupervised uSCA) (ablation) | SCA · USPS+MNIST and MSRC+VOC2007 (1NN, avg) | accuracy (%): 48.12 -> 48.70 | Improved |
| add label information (SCA vs uSCA) (ablation) | SCA · MSRC+VOC2007 | accuracy (%): 33.14 vs 32.75 | No clear effect |
| add label information (SCA vs uSCA) (ablation) | SCA · USPS to MNIST | accuracy (%): 44.86 -> 48.00 | Improved |
| add label information (SCA vs uSCA) (ablation) | SCA · Office+Caltech SURF-BoW (1NN, avg of 12) | accuracy (%): 42.28 -> 44.29 | Improved |
| add label information (SCA vs uSCA) (ablation) | SCA · Office+Caltech DeCAF6 (1NN, avg of 12) | accuracy (%): 82.75 -> 85.88 | Improved |
| use SCA instead of TJM | SCA · Office+Caltech DeCAF6 (1NN, avg of 12) | accuracy (%): 85.59 -> 85.88 | Improved |
| use SCA instead of TJM | SCA · Office+Caltech SURF-BoW (1NN, avg of 12) | accuracy (%): 43.67 -> 44.29 | Improved |
| use SCA instead of TJM | SCA · USPS+MNIST and MSRC+VOC2007 (1NN, avg) | accuracy (%): 48.37 -> 48.70 | Improved |
| use DeCAF6 instead of SURF-BoW features | Raw 1NN · Office+Caltech | accuracy (%): 31.37 -> 70.38 | Improved |
| use SCA instead of TJM for speed | SCA · MNIST+USPS runtime | accuracy (%): 269.44 -> 42.74 | Improved |
| use SCA instead of TSC for speed | SCA · MNIST+USPS runtime | accuracy (%): 3072.25 -> 42.74 | Improved |
| replace 1NN with L-SVM on raw features | Raw features · USPS+MNIST and MSRC+VOC2007 | accuracy (%): 41.89 -> 38.23 | Hurt |
Closest papers by meaning
- Domain Adaptation by Mixture of Alignments of Second- or Higher-Order Scatter Tensors (CVPR 2017 · similarity 0.89)
- SoMA: Singular Value Decomposed Minor Components Adaptation for Domain Generalizable Representation Learning (CVPR 2025 · similarity 0.88)
- Semi-Supervised Heterogeneous Domain Adaptation: Theory and Algorithms (TPAMI 2022 · similarity 0.87)
- Forget More to Learn More: Domain-specific Feature Unlearning for Semi-supervised and Unsupervised Domain Adaptation (ECCV 2024 · similarity 0.87)
- LEARNING FACTORIZED REPRESENTATIONS FOR OPEN-SET DOMAIN ADAPTATION (ICLR 2019 · similarity 0.87)
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 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.
