2022/01/28 by Shilong Liu, Feng Li, Liu, Shilong +13 · 95 citations
Computer Science · #Advanced Neural Network Applications #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #cs.CV
paper · pdf · doi:10.48550/arxiv.2201.12329
Accepted to ICLR 2022
openalex publication_date 2022/01/28 · arxiv created 2022/03/30 · arxiv updated 2022/03/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
We present in this paper a novel query formulation using dynamic anchor boxes for DETR (DEtection TRansformer) and offer a deeper understanding of the role of queries in DETR. This new formulation directly uses box coordinates as queries in Transformer decoders and dynamically updates them layer-by-layer. Using box coordinates not only helps using explicit positional priors to improve the query-to-feature similarity and eliminate the slow training convergence issue in DETR, but also allows us to modulate the positional attention map using the box width and height information. Such a design makes it clear that queries in DETR can be implemented as performing soft ROI pooling layer-by-layer in a cascade manner. As a result, it leads to the best performance on MS-COCO benchmark among the DETR-like detection models under the same setting, e.g., AP 45.7% using ResNet50-DC5 as backbone trained in 50 epochs. We also conducted extensive experiments to confirm our analysis and verify the effectiveness of our methods. Code is available at \urlhttps://github.com/SlongLiu/DAB-DETR.