Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning Transferrable Representations for Unsupervised Domain Adaptation
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 |
|---|---|---|---|
| Joint transductive labelling with reject option (full method) versus k-NN only (ablation) | domain adaptation model · MNIST→SVHN | accuracy: 0.158 -> 0.403 | Helped |
| Adding the reject option to the no-reject variant (ablation) | domain adaptation model · MNIST→SVHN | accuracy: 0.323 -> 0.403 | Helped |
| Adding the reject option to the no-reject variant (ablation) | domain adaptation model · SVHN→MNIST | accuracy: 0.774 -> 0.788 | Helped |
| Adding the reject option to the no-reject variant (ablation) | domain adaptation model · MNIST→MNIST-M | accuracy: 0.835 -> 0.839 | Helped |
| Adding joint labelling and reject option to k-NN-only variant (ablation) | domain adaptation model · Office Webcam→Amazon | accuracy: 0.727 -> 0.811 | Helped |
| Joint labelling and reject option versus k-NN only (no reject option effect on SVHN→MNIST and MNIST→MNIST-M, k-NN-only is worse) (ablation) | domain adaptation model · SVHN→MNIST | accuracy: k-NN only .713 vs full .788 (SVHN→MNIST); k-NN only .158 vs full .403 (MNIST→SVHN) | Mixed |
| Adding joint labelling and reject option to k-NN-only variant (ablation) | domain adaptation model · Office Webcam→D-SLR | accuracy: 0.915 -> 0.992 | Helped |
| Full method (learned metric with transduction) versus BACKPROP baseline | domain adaptation model · Office Amazon→Webcam and D-SLR→Amazon, MNIST→SVHN | accuracy: Amazon→Webcam .536 -> .638; D-SLR→Amazon .728 -> .841; MNIST→SVHN .289 -> .403 | Helped |
| Full method versus BACKPROP on Office Webcam→D-SLR and D-SLR→Webcam | domain adaptation model · Office Webcam→D-SLR, D-SLR→Webcam | accuracy: .992 vs .992; .964 vs .964 | No clear effect |
Closest papers by meaning
- Unsupervised Domain Adaptation with Residual Transfer Networks (NeurIPS 2016 · similarity 0.92)
- Unsupervised Domain Adaptation With Similarity Learning (CVPR 2018 · similarity 0.92)
- Unsupervised Domain Adaptation by Matching Distributions Based on the Maximum Mean Discrepancy via Unilateral Transformations (AAAI 2019 · similarity 0.91)
- Co-regularized Alignment for Unsupervised Domain Adaptation (NeurIPS 2018 · similarity 0.91)
- A General Representation-Based Approach to Multi-Source Domain Adaptation (ICML 2025 · similarity 0.90)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · similarity 0.89)
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.