Abstract:Cross-Domain Few-Shot Learning (CD-FSL) is a recently emerging task that tackles few-shot learning across different domains. It aims at transferring prior knowledge learned on the source dataset to novel target datasets. The CD-FSL task is especially challenged by the huge domain gap between different datasets. Critically, such a domain gap actually comes from the changes of visual styles, and wave-SAN empirically shows that spanning the style distribution of the source data helps alleviate this issue. However, wave-SAN simply swaps styles of two images. Such a vanilla operation makes the generated styles ``real'' and ``easy'', which still fall into the original set of the source styles. Thus, inspired by vanilla adversarial learning, a novel model-agnostic meta Style Adversarial training (StyleAdv) method together with a novel style adversarial attack method is proposed for CD-FSL. Particularly, our style attack method synthesizes both ``virtual'' and ``hard'' adversarial styles for model training. This is achieved by perturbing the original style with the signed style gradients. By continually attacking styles and forcing the model to recognize these challenging adversarial styles, our model is gradually robust to the visual styles, thus boosting the generalization ability for novel target datasets. Besides the typical CNN-based backbone, we also employ our StyleAdv method on large-scale pretrained vision transformer. Extensive experiments conducted on eight various target datasets show the effectiveness of our method. Whether built upon ResNet or ViT, we achieve the new state of the art for CD-FSL. Code is available at <a class="link-external link-https" href="https://github.com/lovelyqian/StyleAdv-CDFSL" rel="external noopener nofollow">this https URL</a>.

Robust Object Detection Via Adversarial Novel Style Exploration.

Domain Adaptation for Object Detection via Style Consistency

Towards Improving Robustness Against Common Corruptions in Object Detectors Using Adversarial Contrastive Learning

Robust and Accurate Object Detection via Adversarial Learning

Toward Enhanced Adversarial Robustness Generalization in Object Detection: Feature Disentangled Domain Adaptation for Adversarial Training

Adversarial Example Generation Method for Object Detection in Remote Sensing Images

Adversarial Style Augmentation for Domain Generalized Urban-Scene Segmentation

Style-Guided Adversarial Teacher for Cross-Domain Object Detection

AdvLogo: Adversarial Patch Attack against Object Detectors based on Diffusion Models

StyleAdv: Meta Style Adversarial Training for Cross-Domain Few-Shot Learning

Robust Remote Sensing Scene Interpretation Based on Unsupervised Domain Adaptation

Strong-Weak Distribution Alignment for Adaptive Object Detection

Uncertainty-Aware Unsupervised Domain Adaptation in Object Detection

DA-RAW: Domain Adaptive Object Detection for Real-World Adverse Weather Conditions

Adapting Object Detectors with Conditional Domain Normalization

Region Feature Disentanglement for Domain Adaptive Object Detection

A Step-Wise Domain Adaptation Detection Transformer for Object Detection under Poor Visibility Conditions

Adversarial Defense Teacher for Cross-Domain Object Detection under Poor Visibility Conditions

AWADA: Attention-Weighted Adversarial Domain Adaptation for Object Detection

Toward Improving Robustness of Object Detectors Against Domain Shift

SDNIA-YOLO: A Robust Object Detection Model for Extreme Weather Conditions