Research map / Label-efficient and robust learning / Domain adaptation and generalization
Cross-Domain Cross-Set Few-Shot Learning via Learning Compact and Aligned Representations
Wentao Chen, Zhang Zhang, Wei Wang, Liang Wang, Zilei Wang, Tieniu Tan
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 strong augmentation to source-only training (ablation) | stabPA · DomainNet real-sketch 1-shot | accuracy: 40.23 -> 41.74 | Improved |
| add source-to-target alignment (ablation) | stabPA · DomainNet real-sketch 1-shot | accuracy: 40.23 -> 42.86 | Improved |
| add target-to-source alignment (ablation) | stabPA · DomainNet real-sketch 1-shot | accuracy: 40.23 -> 44.20 | Improved |
| combine both alignments (ablation) | stabPA · DomainNet real-sketch 1-shot | accuracy: 44.20 -> 47.01 | Improved |
| add strong augmentation to bi-directional alignment (ablation) | stabPA · DomainNet real-sketch 1-shot | accuracy: 47.01 -> 50.85 | Improved |
| add strong augmentation to bi-directional alignment (ablation) | stabPA · DomainNet real-sketch 5-shot | accuracy: 56.68 -> 61.37 | Improved |
| add strong augmentation to bi-directional alignment (ablation) | stabPA · DomainNet sketch-real 5-shot | accuracy: 64.32 -> 68.93 | Improved |
| use 100% instead of 10% of unlabeled samples (ablation) | stabPA · DomainNet avg 1-shot | accuracy: 51.76 -> 53.92 | Improved |
| use 10% unlabeled samples vs FixMatch with 100% | stabPA · DomainNet avg 5-shot | accuracy: 62.58 -> 65.96 | Improved |
| include unlabeled data from base classes (0% to 100% of classes) (ablation) | stabPA · DomainNet avg 1-shot | accuracy: 50.74 -> 53.92 | Improved |
| set interpolation coefficient lambda 0.2 instead of 0.0 (ablation) | stabPA · DomainNet real-painting 5-shot | accuracy: 64.95 -> 65.65 | Improved |
| raise interpolation coefficient lambda from 0.2 to 1.0 (ablation) | stabPA · DomainNet real-painting 5-shot | accuracy: 65.65 -> 60.60 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Dual Adaptive Representation Alignment for Cross-Domain Few-Shot Learning (TPAMI 2023 · similarity 0.92)
- Understanding Cross-Domain Few-Shot Learning Based on Domain Similarity and Few-Shot Difficulty (NeurIPS 2022 · similarity 0.89)
- MMT: Cross Domain Few-Shot Learning via Meta-Memory Transfer (TPAMI 2023 · similarity 0.89)
- Cross-Domain Few-Shot Segmentation via Iterative Support-Query Correspondence Mining (CVPR 2024 · similarity 0.87)
- Discriminative Sample-Guided and Parameter-Efficient Feature Space Adaptation for Cross-Domain Few-Shot Learning (CVPR 2024 · similarity 0.87)
- Flatten Long-Range Loss Landscapes for Cross-Domain Few-Shot Learning (CVPR 2024 · 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.
