[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83918-en":3,"doc-seo-83918-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},83918,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","FSDC-DETR Frequency-Spatial Domain Collaborative DETR for Small Object Detection","Small object detection (SOD) remains difficult in real-world scenarios because conventional detectors entangle spatial aggregation with frequency aliasing and truncation, degrading high-frequency information needed for tiny targets. FSDC-DETR introduces a Frequency-Spatial Domain Collaborative Detection Transformer that explicitly models complementary spatial and frequency representations. Dual-Branch Frequency-Spatial Adaptive Fusion and Shunt Frequency-Spatial Feature Fusion enable bidirectional interaction and progressive cross-scale propagation, while Frequency-Spatial Dynamic Downsampling preserves informative high-frequency responses during scale transitions. Experiments show state-of-the-art gains, including +6.4 AP on VisDrone-DET2019 and +6.6 on AITODv2.","arXiv :2607 .05 176v2 [ cs .CV] 7 Jul 2026  \nFSDC-DETR: A Frequency-Spatial Domain Collaborative DETR for Small Object Detection  \nAiwen Liu 1 , Chengguang Zhu 1 (\\#), Gang Wang 1 , Dandan Zhu2 , Haodong Lin 1 , Yan Wang4 , Huiyu Zhou3 , and Zhengyi Pan 1  \n1 Micro-Intelligence, Shanghai 201100, China  \n[nevereverinsomnia@gmail.com](nevereverinsomnia@gmail.com) , [chengguang.zhusjtu@gmail.com](chengguang.zhusjtu@gmail.com) ,{[roy.wang](roy.wang),haodong.lin,[eric.pan}@micro-i.com.cn](eric.pan}@micro-i.com.cn)  \n2 East China Normal University, Shanghai 200241, China  \n[ddzhu@mail.ecnu.edu.cn](ddzhu@mail.ecnu.edu.cn)  \n3 University of Leicester, Leicester LE1 7RH, UK  \n[hz143@leicester.ac.uk](hz143@leicester.ac.uk)  \n4 Chongqing Normal University, Chongqing 401331, China  \nAbstract. Small object detection (SOD) remains a challenging task in real-world applications. Despite recent advances, existing detectors remain limited by rigid processing that entangle spatial aggregation with implicit frequency aliasing and truncation, leading to inadequate preservation of high-frequency components for SOD. To tackle these limitations, we propose a Frequency-Spatial Domain Collaborative Detection Transformer (FSDC-DETR), a novel collaborative framework that explicitly models complementary spatial and frequency representations.  \nSpecifically, we first introduce Dual-Branch Frequency-Spatial Adaptive Fusion (DBFSAF) to enhance frequency diversity and adaptively capture frequency-spatial domain discriminative representations. Building on these representations, a frequency-spatial interaction scheme is further explored within the hybrid encoder to enable progressive feature propagation to the decoder. In particular, structure-aware frequencyspatial aggregation is achieved through Shunt Frequency-Spatial Feature Fusion (SFS-FF), establishing bidirectional interaction and progressive cross-scale propagation between frequency and spatial representations for coherent discriminative modeling. Meanwhile, informative high-frequency responses are preserved during scale transitions through Frequency-Spatial Dynamic Downsampling (FSD-Down), thereby minimizing frequency degradation throughout multi-scale fusion for the precise SOD. Experimental results demonstrate that FSDC-DETR achieves state-of-the-art performance, improving AP by 6.4 on VisDrone-DET2019 and 6.6 on AITODv2, with gains of 6.8 and 6.9 AP for small objects. The code is available at [https://github.com/nevereverinsomnia/FSDC](https://github.com/nevereverinsomnia/FSDC)DETR.  \nKeywords: small object detection · detection transformer · frequencyspatial collaborative modeling · multi-scale feature fusion  \n⋆ Aiwen Liu, Chengguang Zhu, and Gang Wang—Equal contribution.  \n2 Liu et al.  \n1 Introduction  \nObject detection is a fundamental task in computer vision that has witnessed significant progress in recent years, supporting a wide range of applications such as industrial defect detection [12, 37, 79, 90], aerial surveillance [11, 22, 39, 80], remote sensing [2,34,53,54], medical lesion analysis [16,73,88], and autonomous systems [13, 52, 75, 83] . Most of these advances have been driven by Convolutional Neural Network (CNN)-based detectors, which typically rely on a series of manually designed components, such as anchor box generation and nonmaximum suppression (NMS), and remain heavily dependent on human expertise [18,27,49,63] . These components often require extensive task-specific tuning, making the overall detection pipeline complex and inefficient in real-world applications. More recently, the end-to-end Transformer-based Detection Transformer (DETR) framework has been introduced to largely alleviate these limitations by reformulating object detection as a direct set prediction problem [4] . Subsequent DETR-based variants have further improved the detection accuracy while significantly improving inference efficiency [25, 35, 43, 92] . However, these advances are primarily designed for g","cbCaidZiFUbxJF2j","https://ap.wps.com/l/cbCaidZiFUbxJF2j","pdf",11616191,1,23,"English","en",105,"# Introduction\n# Frequency Response Analysis\n# Proposed Method: FSDC-DETR\n## Dual-Branch Frequency-Spatial Adaptive Fusion (DBFSAF)\n## Shunt Frequency-Spatial Feature Fusion (SFS-FF)\n## Frequency-Spatial Dynamic Downsampling (FSD-Down)\n# Experiments and Results\n# Conclusion","[{\"question\":\"Why do existing detectors underperform on small object detection?\",\"answer\":\"They often combine spatial aggregation with implicit frequency aliasing and truncation, which inadequately preserves high-frequency components crucial for tiny targets.\"},{\"question\":\"What is the core idea behind FSDC-DETR?\",\"answer\":\"FSDC-DETR explicitly models complementary spatial and frequency representations through a frequency-spatial domain collaborative framework.\"},{\"question\":\"How does FSDC-DETR maintain high-frequency information across multi-scale fusion?\",\"answer\":\"It uses Frequency-Spatial Dynamic Downsampling (FSD-Down) to preserve informative high-frequency responses during scale transitions, minimizing frequency degradation.\"}]",1784191433,58,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"fsdc-detr-frequency-spatial-domain-collaborative-detr-for-small-object-detection","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/fsdc-detr-frequency-spatial-domain-collaborative-detr-for-small-object-detection/83918/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"Why do existing detectors underperform on small object detection?","Question",{"text":74,"@type":75},"They often combine spatial aggregation with implicit frequency aliasing and truncation, which inadequately preserves high-frequency components crucial for tiny targets.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What is the core idea behind FSDC-DETR?",{"text":79,"@type":75},"FSDC-DETR explicitly models complementary spatial and frequency representations through a frequency-spatial domain collaborative framework.",{"name":81,"@type":72,"acceptedAnswer":82},"How does FSDC-DETR maintain high-frequency information across multi-scale fusion?",{"text":83,"@type":75},"It uses Frequency-Spatial Dynamic Downsampling (FSD-Down) to preserve informative high-frequency responses during scale transitions, minimizing frequency degradation.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]