Research map / Label-efficient and robust learning / Domain adaptation and generalization
Language Does Matter for Cross-Domain Few-Shot Visual Feature Enhancement
Fei Zhou, Xiwen Zhang, Qingqing Qiu, Lei Zhang, Wei Wei, Chen Ding, Yi Zhang, Liang Li, Xiangyu Yue, Yanning 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 |
|---|---|---|---|
| add language-based feature refinement to StyleAdv | StyleAdv (ResNet-10) · CD-FSC average, 1-shot | accuracy (%): 53.00 -> 55.50 | Improved |
| add language-based feature refinement to PMF | PMF (ViT-Small) · CD-FSC average, 1-shot | accuracy (%): 50.91 -> 55.74 | Improved |
| add language-based feature refinement to CD-CLS | CD-CLS (ViT-Small) · CD-FSC average, 1-shot | accuracy (%): 54.62 -> 57.98 | Improved |
| add language-based feature refinement to IFA | IFA (ResNet-50) · CD-FSS average, 1-shot | mIoU: 67.8 -> 71.9 | Improved |
| add language-based feature refinement to CD-ViTO | CD-ViTO (ViT-Large) · CD-FSOD average, 10-shot | mAP: 29.6 -> 36.9 | Improved |
| add image-level attributes only (ablation) | IFA (ResNet-50) · CD-FSS ISIC, 1-shot | mIoU: 66.3 -> 70.6 | Improved |
| add domain-level attributes only (ablation) | IFA (ResNet-50) · CD-FSS ISIC, 1-shot | mIoU: 66.3 -> 70.1 | Improved |
| combine image-level and domain-level attributes (ablation) | IFA (ResNet-50) · CD-FSS ISIC, 1-shot | mIoU: 70.6 -> 73.3 | Improved |
| add four extra MLP layers without language to baseline (ablation) | IFA (ResNet-50) · CD-FSS ISIC, 1-shot | mIoU: 66.3 -> 65.7 | Hurt |
| add residual connection in the feature refiner (ablation) | Ours · CD-FSS ISIC, 1-shot | mIoU: 64.8 -> 73.3 | Improved |
| use cross-attention instead of element-wise addition (ablation) | Ours · CD-FSS ISIC, 1-shot | mIoU: 65.7 -> 73.3 | Improved |
| use cross-attention instead of element-wise multiplication (ablation) | Ours · CD-FSS ISIC, 1-shot | mIoU: 70.7 -> 73.3 | Improved |
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.
- StyleAdv — StyleAdv: Meta Style Adversarial Training for Cross-Domain Few-Shot Learning (CVPR 2023) · 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
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.88)
- The Devil is in Low-Level Features for Cross-Domain Few-Shot Segmentation (CVPR 2025 · similarity 0.87)
- From Local Details to Global Context: Advancing Vision-Language Models with Attention-Based Selection (ICML 2025 · similarity 0.87)
- Mind the Gap Between Prototypes and Images in Cross-domain Finetuning (NeurIPS 2024 · similarity 0.87)
- Adapt Before Comparison: A New Perspective on Cross-Domain Few-Shot Segmentation (CVPR 2024 · similarity 0.86)
- UMFC: Unsupervised Multi-Domain Feature Calibration for Vision-Language Models (NeurIPS 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.
