{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T12:59:45Z","timestamp":1740142785213,"version":"3.37.3"},"reference-count":42,"publisher":"Oxford University Press (OUP)","issue":"12","license":[{"start":{"date-parts":[[2024,9,16]],"date-time":"2024-09-16T00:00:00Z","timestamp":1726444800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/pages\/standard-publication-reuse-rights"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62372125","62372164"],"award-info":[{"award-number":["62372125","62372164"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003392","name":"Natural Science Foundation of Fujian Province","doi-asserted-by":"publisher","award":["2023J01246"],"award-info":[{"award-number":["2023J01246"]}],"id":[{"id":"10.13039\/501100003392","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Guangdong Natural Science Funds for Distinguished Young Scholar","award":["2023B1515020041"],"award-info":[{"award-number":["2023B1515020041"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2024,12,20]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Sparse R-CNN is a new paradigm of object detection, which predicts objects in a sparse way. However, there are some limitations in Sparse R-CNN. One is the presence of weak prior information caused by fixed learnable proposal boxes and features across different images, necessitating excessive iterations for the model to refine its predictions; the other is the inadequate exploitation of multi-scale information, leading to the sub-optimal detection performance. Thus, building upon Sparse R-CNN, we propose an efficient detector that incorporates dynamic prior and dynamic feature fusion, called $D^{2}$-Det. In particular, for the dynamic prior part, a prior information generator module dynamically generates proposal features and boxes as the dynamic prior for different images to alleviate the inference-inefficient iterative refinement process of predictions, and we further propose the class scores decoupling method to reduce the computation overhead. Furthermore, for the dynamic feature fusion part, we develop a novel lightweight multi-scale feature fusion module, which dynamically aggregates features from all layers for each proposal box, enabling adaptive feature fusion and improving detection precision by nearly 2 AP. Experiments show that $D^{2}$-Det can achieve 46.6 AP on COCO 2017 with fewer computations for the backbone ResNet50, surpassing most of the state-of-the-art detectors.<\/jats:p>","DOI":"10.1093\/comjnl\/bxae082","type":"journal-article","created":{"date-parts":[[2024,9,17]],"date-time":"2024-09-17T17:04:27Z","timestamp":1726592667000},"page":"3196-3206","source":"Crossref","is-referenced-by-count":0,"title":["Efficient object detector via dynamic prior and dynamic feature fusion"],"prefix":"10.1093","volume":"67","author":[{"given":"Zihang","family":"Zhang","sequence":"first","affiliation":[{"name":"College of Computer Science and Electronic Engineering , Hunan University, Changsha 410082,","place":["China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuling","family":"Liu","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering , Hunan University, Changsha 410082,","place":["China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhili","family":"Zhou","sequence":"additional","affiliation":[{"name":"Institute of Artificial Intelligence and Blockchain , Guangzhou University, Guangzhou 510006,","place":["China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Gaobo","family":"Yang","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering , Hunan University, Changsha 410082,","place":["China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Q M Jonathan","family":"Wu","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering , University of Windsor, Windsor 250101,","place":["Canada"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"286","published-online":{"date-parts":[[2024,9,16]]},"reference":[{"key":"2025010523430497700_ref1","first-page":"580","article-title":"Rich feature hierarchies for accurate object detection and semantic segmentation","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, June 23\u201328","author":"Girshick","year":"2014"},{"key":"2025010523430497700_ref2","first-page":"1440","article-title":"Fast R-CNN","volume-title":"Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile, December 7\u201313","author":"Girshick","year":"2015"},{"key":"2025010523430497700_ref3","doi-asserted-by":"publisher","first-page":"154","DOI":"10.1007\/s11263-013-0620-5","article-title":"Selective search for object recognition","volume":"104","author":"Uijlings","year":"2013","journal-title":"Int. J. Comput. Vis."},{"key":"2025010523430497700_ref4","doi-asserted-by":"publisher","first-page":"1137","DOI":"10.1109\/TPAMI.2016.2577031","article-title":"Faster R-CNN: Towards real-time object detection with region proposal networks","volume":"39","author":"Ren","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2025010523430497700_ref5","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"In Advances in Neural Information Processing Systems, Long Beach, CA, USA, December"},{"key":"2025010523430497700_ref6","first-page":"213","article-title":"End-to-end object detection with transformers","volume-title":"Proceedings of the European Conference on Computer Vision, Glasgow, UK, august 23-28","author":"Carion","year":"2020"},{"author":"Yao","key":"2025010523430497700_ref7","article-title":"Efficient DETR: improving end-to-end object detector with dense prior."},{"key":"2025010523430497700_ref8","article-title":"Deformable DETR: Deformable transformers for end-to-end object detection","volume-title":"Proc. of 9th int. conf. On learning representations, virtual event, Austria, may 3-7","author":"Zhu","year":"2021"},{"key":"2025010523430497700_ref9","first-page":"2567","article-title":"Anchor DETR: query design for transformer-based detector","volume-title":"Proceedings of Thirty-Sixth AAAI Conference on Artificial Intelligence, Virtual Event, February 22 - March 1","author":"Wang","year":"2022"},{"key":"2025010523430497700_ref10","first-page":"3601","article-title":"Fast convergence of DETR with spatially modulated co-attention","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision, Montreal, QC, Canada, October 10-17","author":"Gao","year":"2021"},{"key":"2025010523430497700_ref11","first-page":"13609","article-title":"DN-DETR: Accelerate DETR training by introducing query denoising","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, June 18-24","author":"Li","year":"2022"},{"key":"2025010523430497700_ref12","first-page":"3631","article-title":"Conditional DETR for fast training convergence","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision, Montreal, QC, Canada, October 10-17","author":"Meng","year":"2021"},{"key":"2025010523430497700_ref13","first-page":"3591","article-title":"Rethinking transformer-based set prediction for object detection","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision, Montreal, QC, Canada, October 10-17","author":"Sun","year":"2021"},{"key":"2025010523430497700_ref14","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR46437.2021.01422","article-title":"Sparse R-CNN: end-to-end object detection with learnable proposals","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 19-25.","author":"Sun","year":"2021"},{"key":"2025010523430497700_ref15","first-page":"332","article-title":"You should look at all objects","volume-title":"Proceedings of the European Conference on Computer Vision, Tel Aviv, Israel, October 23-27","author":"Jin","year":"2022"},{"key":"2025010523430497700_ref16","first-page":"936","article-title":"Feature pyramid networks for object detection","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, July 21-26","author":"Lin","year":"2017"},{"key":"2025010523430497700_ref17","doi-asserted-by":"publisher","first-page":"1904","DOI":"10.1109\/TPAMI.2015.2389824","article-title":"Spatial pyramid pooling in deep convolutional networks for visual recognition","volume":"37","author":"He","year":"2015","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"2025010523430497700_ref18","first-page":"379","article-title":"R-FCN: Object detection via region-based fully convolutional networks","volume-title":"Advances in Neural Information Processing Systems, Barcelona, Spain, December 5-10","author":"Dai","year":"2016"},{"key":"2025010523430497700_ref19","first-page":"2980","article-title":"Mask R-CNN","volume-title":"Proceedings of the IEEE International Conference On computer vision, Venice, Italy, October 22-29","author":"He","year":"2017"},{"key":"2025010523430497700_ref20","first-page":"779","article-title":"You Only Look Once: unified, real-time object detection","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, June 27-30","author":"Redmon","year":"2016"},{"key":"2025010523430497700_ref21","first-page":"21","article-title":"SSD: Single Shot MultiBox Detector","volume-title":"Proceedings of the European Conference on Computer Vision, Amsterdam, the Netherlands, October 11-14","author":"Liu","year":"2016"},{"key":"2025010523430497700_ref22","first-page":"2999","article-title":"Focal loss for dense object detection","volume-title":"Proceedings of the IEEE International Conference On computer vision, Venice, Italy, October 22-29","author":"Lin","year":"2017"},{"key":"2025010523430497700_ref23","first-page":"6517","article-title":"YOLO9000: Better, faster, stronger","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, July 21-26","author":"Redmon","year":"2017"},{"author":"Redmon","key":"2025010523430497700_ref24","article-title":"YOLOv3: an incremental improvement."},{"author":"Bochkovskiy","key":"2025010523430497700_ref25","article-title":"YOLOv4: optimal speed and accuracy of object"},{"author":"Ge","key":"2025010523430497700_ref26","article-title":"YOLOX: exceeding YOLO series in 2021."},{"key":"2025010523430497700_ref27","first-page":"9626","article-title":"FCOS: Fully Convolutional One-Stage Object Detection","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision, Seoul, Korea (south), October 27 - November 2","author":"Tian","year":"2019"},{"key":"2025010523430497700_ref28","first-page":"765","article-title":"Cornernet: Detecting objects as paired keypoints","volume-title":"Proceedings of the European Conference on Computer Vision, Munich, Germany, September 8-14","author":"Law","year":"2018"},{"author":"Zhu","key":"2025010523430497700_ref29","article-title":"AutoAssign: differentiable label assignment for dense object detection."},{"key":"2025010523430497700_ref30","first-page":"4641","article-title":"PNP-DETR: towards efficient visual analysis with transformers","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision, Montreal, QC, Canada, October 10-17","author":"Wang","year":"2021"},{"key":"2025010523430497700_ref31","article-title":"Sparse DETR: Efficient end-to-end object detection with learnable sparsity","volume-title":"Proceedings of Tenth International Conference on Learning Representations, Virtual Event, April 25-29","author":"Roh","year":"2022"},{"key":"2025010523430497700_ref32","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2018.00913","article-title":"Path aggregation network for instance segmentation","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, June 18-22","author":"Liu","year":"2018"},{"author":"Chen","key":"2025010523430497700_ref33","article-title":"Feature pyramid grids."},{"key":"2025010523430497700_ref34","doi-asserted-by":"publisher","first-page":"1028","DOI":"10.1093\/comjnl\/bxab055","article-title":"LEDet: a single-shot real-time object detector based on low-light image enhancement","volume":"64","author":"Hao","year":"2021","journal-title":"Comput. J."},{"key":"2025010523430497700_ref35","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR42600.2020.01079","article-title":"Efficientdet: Scalable and efficient object detection","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, June 13-19","author":"Tan","year":"2020"},{"key":"2025010523430497700_ref36","doi-asserted-by":"publisher","first-page":"146","DOI":"10.1016\/j.neucom.2021.04.002","article-title":"Adaptive multiscale feature for object detection","volume":"449","author":"Yu","year":"2021","journal-title":"Neurocomputing"},{"key":"2025010523430497700_ref37","first-page":"770","article-title":"Deep residual learning for image recognition","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, June 27-30","author":"He","year":"2016"},{"key":"2025010523430497700_ref38","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-10602-1_48","article-title":"Microsoft COCO: Common Objects in Context","volume-title":"Proceedings of the European Conference on Computer Vision, Zurich, Switzerland, September 6-12","author":"Lin","year":"2014"},{"key":"2025010523430497700_ref39","article-title":"Decoupled weight decay regularization","volume-title":"Proceedings of 7th International Conference on Learning Representations, New Orleans, LA, USA, May 6\u20139","author":"Loshchilov","year":"2019"},{"key":"2025010523430497700_ref40","first-page":"1106","article-title":"Imagenet classification with deep convolutional neural networks","volume-title":"Advances in Neural Information Processing Systems, Lake Tahoe, Nevada, United States, December 3-6","author":"Krizhevsky","year":"2012"},{"key":"2025010523430497700_ref41","doi-asserted-by":"publisher","first-page":"83","DOI":"10.1002\/nav.3800020109","article-title":"The Hungarian method for the assignment problem","volume":"2","author":"Kuhn","year":"1955","journal-title":"Naval Res. Logist. Q."},{"key":"2025010523430497700_ref42","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2019.00075","article-title":"Generalized intersection over union: a metric and a loss for bounding box regression","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA, June 16-20","author":"Rezatofighi","year":"2019"}],"container-title":["The Computer Journal"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/comjnl\/article-pdf\/67\/12\/3196\/59140975\/bxae082.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/comjnl\/article-pdf\/67\/12\/3196\/59140975\/bxae082.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,1,6]],"date-time":"2025-01-06T04:32:26Z","timestamp":1736137946000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/comjnl\/article\/67\/12\/3196\/7758619"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,9,16]]},"references-count":42,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2024,9,16]]},"published-print":{"date-parts":[[2024,12,20]]}},"URL":"https:\/\/doi.org\/10.1093\/comjnl\/bxae082","relation":{},"ISSN":["0010-4620","1460-2067"],"issn-type":[{"type":"print","value":"0010-4620"},{"type":"electronic","value":"1460-2067"}],"subject":[],"published-other":{"date-parts":[[2024,12]]},"published":{"date-parts":[[2024,9,16]]}}}