DG-DETR: Toward Domain Generalized Detection Transformer

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hwang, Seongmin, Han, Daeyoung, Jeon, Moongu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909898903650304
author Hwang, Seongmin
Han, Daeyoung
Jeon, Moongu
author_facet Hwang, Seongmin
Han, Daeyoung
Jeon, Moongu
contents 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 at https://github.com/sminhwang/DG-DETR.
format Preprint
id arxiv_https___arxiv_org_abs_2504_19574
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DG-DETR: Toward Domain Generalized Detection Transformer
Hwang, Seongmin
Han, Daeyoung
Jeon, Moongu
Computer Vision and Pattern Recognition
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 at https://github.com/sminhwang/DG-DETR.
title DG-DETR: Toward Domain Generalized Detection Transformer
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.19574