DG-DETR: Toward Domain Generalized Detection Transformer

End-to-end Transformer-based detectors (DETRs) have demonstrated strong detection performance. However, domain generalization (DG) research has primarily focused on convolutional neural network (CNN)-based detectors, while paying little attention to enhancing the robustness of DETRs. In this letter, we introduce a Domain Generalized DEtection TRansformer (DG-DETR), a simple, effective, and plug-and-play method that improves out-of-distribution (OOD) robustness for DETRs. Specifically, we propose a novel domain-agnostic query selection strategy that removes domain-induced biases from object queries via orthogonal projection onto the instance-specific style space. Additionally, we leverage a wavelet decomposition to disentangle features into domain-invariant and domain-specific components, enabling synthesis of diverse latent styles while preserving the semantic features of objects. Experimental results validate the effectiveness of DG-DETR. Our code is available atthis https URL.

@article{hwang2025_2504.19574,
  title={ DG-DETR: Toward Domain Generalized Detection Transformer },
  author={ Seongmin Hwang and Daeyoung Han and Moongu Jeon },
  journal={arXiv preprint arXiv:2504.19574},
  year={ 2025 }
}