[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82910-en":3,"doc-seo-82910-105":30,"detail-sidebar-cat-0-en-105":91},{"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":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},82910,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Fully Rotation-Equivariant Spectral-Spatial Learning for Multispectral Object Detection","Existing multispectral detectors face three core limits: discrete spectral processing, a scale-dependent change in how reliable spectral versus spatial cues are across feature pyramids, and the absence of explicit rotation-equivariant geometric priors for arbitrarily oriented objects. FressDet addresses these issues with a fully rotation-equivariant spectral–spatial learning framework. It combines Spectral Implicit Warp for monotone order-preserving spectral resampling, Rotation-Equivariant Consistency Weighting for reliability-aware fusion across pyramid levels, and an oriented-aware head using group-indexed features for stable oriented predictions. The approach delivers state-of-the-art accuracy with 93% fewer parameters on three public benchmarks, demonstrating robustness under rotation perturbations and strong generalizability.","arXiv :2607 .05 148v 1 [ cs .CV] 6 Jul 2026  \nFully Rotation-Equivariant Spectral-Spatial Learning for Multispectral Object Detection  \nPeng Zhang, Tingfa Xu†, Shuaihao Han, and Jianan Li† Beijing Institute of Technology, Beijing, China  \nAbstract. Existing multispectral detectors are limited by discrete spectral processing, a scale-dependent shift in the relative reliability of spectral and spatial cues across pyramid levels, and the lack of explicit rotation-equivariant geometric priors for arbitrarily oriented objects. To tackle these limitations, we propose FressDet, a fully rotation-equivariant spectral–spatial learning framework for multispectral object detection, capable of capturing the continuous, ordered nature of spectral structure and enabling reliable spectral–spatial fusion across pyramid levels under arbitrary in-plane rotations. FressDet integrates three complementary components. Spectral Implicit Warp (SpeIW) enables query-based spectral resampling via a coordinate-conditioned implicit field, yielding a monotone, order-preserving warp. Rotation-Equivariant Consistency Weighting (ReCoW) adaptively fuses spectral and spatial branches based on branch reliability, reinforcing informative cues while suppressing noise across pyramid levels. The oriented-aware head exploits group-indexed features to stably predict oriented objects without parameter replication.  \nTaken together, FressDet learns more discriminative and robust spectral– spatial representations even under rotational perturbations. By achieving state-of-the-art performance with 93% fewer parameters on three public benchmarks, FressDet demonstrates its effectiveness and generalizability.  \nCode is available at [https://github.com/Riiluo/FressDet](https://github.com/Riiluo/FressDet).  \nKeywords: Multispectral object detection · Equivariant neural networks · Implicit spectral warping · Spectral–spatial learning  \n1 Introduction  \nMultispectral object detection (MOD) improves localization and recognition by jointly exploiting spatial structure and spectral cues. Compared with RGB imagery, multispectral images (MSIs) measure spectral responses at sampled wavelengths, yielding per-pixel spectral signatures with strong inter-band correlation. Prior methods either decouple spectral and spatial information through dimensionality reduction (e.g., PCA [36] and band selection [40, 72]) or rely on attention-based modeling to capture spectral–spatial interactions [15] . Although these methods achieve promising performance, two challenges remain in MOD: (i) discrete spectral processing that neglects the continuous and ordered  \n† Correspondence to: Tingfa Xu and Jianan Li.  \n2 P. Zhang et al.  \n(b) Rot.Perturb.  \nYolo26m 64.4%  DCFL 67.5%  MessDet 68.1%  OSSDet 69.0%  FressDet 73.1%  \n300°  \n270°  \n240°  \n0°  \n60°  \n90°  \n120°  \n210°  \n150° 180°  \n330° 30°  \n(c)76 74  \n72  \n70  \n68  \n66  \n64  \n62  \n60  \n0 10 20 30 40 50 60  \n70  \n\n|  | FressD | et (Ours | ) |  |  |  |  |\n| --- | --- | --- | --- | --- | --- | --- | --- |\n|  |  |  |  B | etter P | erform | ance |  |\n|  |  |  |  | OS | SDet |  |  |\n|  |  | MessDet |  |  |  | O-DETR |  |\n|  |  |  | \u003Cbr>R | \u003Cbr>eDet | DCFL\u003Cbr> |  |  |\n|  | 500G | Y | olo11m |  Stri\u003Cbr> | pRCNN | \u003Cbr>O-R | CNN |\n|  | \u003Cbr>300G |  | Yolo26 | \u003Cbr>m |  | RTMDet |  |\n|  |  100G |  |  | S2ADet |  |  |  |\n\nmAP50 (%)  \nNumber of Parameters (M)  \nFig. 1: (a) Transformation consistency in vanilla (V-CNN) and equivariant (Eq-CNN) networks. Te and Tr denote the identity and a rotation, f is a feature extractor, and Z0 , Z1 are features of I0 = Te (I0) and I1 = Tr (I0) . Ideally Tr− 1 (Z1)≈Z0 ; the error map confirms Eq-CNN yields lower transformation error. However, existing methods have not yet unified continuous spectral modeling with rotation-equivariant geometric priors.(b) Detection robustness under rotational perturbations at 1° intervals on MODA. FressDet exhibits the best robustness. (c) mAP50 vs. parameter count on MODA, where circle size is pr","cbCaic3Y8M20WYsd","https://ap.wps.com/l/cbCaic3Y8M20WYsd","pdf",5276492,2,1,20,"English","en",105,"# Introduction\n## Challenges in multispectral object detection\n## Proposed method: FressDet\n## Contributions and evaluation overview","[{\"question\":\"What limitations of existing multispectral detectors does FressDet target?\",\"answer\":\"FressDet targets discrete spectral processing, scale-dependent reliability shifts between spectral and spatial cues across pyramid levels, and the lack of explicit rotation-equivariant geometric priors for arbitrarily oriented objects.\"},{\"question\":\"How does Spectral Implicit Warp (SpeIW) improve spectral modeling?\",\"answer\":\"SpeIW performs query-based spectral resampling using a coordinate-conditioned implicit field, producing a monotone, order-preserving warp that captures continuous spectral structure.\"},{\"question\":\"How does FressDet handle rotation equivariance for oriented object prediction?\",\"answer\":\"FressDet uses a rotation-equivariant design with a reliability-aware fusion module and an oriented-aware head that employs group-indexed features, enabling stable oriented predictions without parameter replication.\"}]",1784183879,50,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"fully-rotation-equivariant-spectral-spatial-learning-for-multispectral-object-detection","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,47,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":20},"https://docshare.wps.com/document/","Document",{"item":48,"name":12,"@type":43,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/fully-rotation-equivariant-spectral-spatial-learning-for-multispectral-object-detection/82910/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-23","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What limitations of existing multispectral detectors does FressDet target?","Question",{"text":75,"@type":76},"FressDet targets discrete spectral processing, scale-dependent reliability shifts between spectral and spatial cues across pyramid levels, and the lack of explicit rotation-equivariant geometric priors for arbitrarily oriented objects.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does Spectral Implicit Warp (SpeIW) improve spectral modeling?",{"text":80,"@type":76},"SpeIW performs query-based spectral resampling using a coordinate-conditioned implicit field, producing a monotone, order-preserving warp that captures continuous spectral structure.",{"name":82,"@type":73,"acceptedAnswer":83},"How does FressDet handle rotation equivariance for oriented object prediction?",{"text":84,"@type":76},"FressDet uses a rotation-equivariant design with a reliability-aware fusion module and an oriented-aware head that employs group-indexed features, enabling stable oriented predictions without parameter replication.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,126,129,133],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":29,"slug":113},6,"Technology","technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":22,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":127,"show_sort_weight":22,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":46,"category_name":135,"show_sort_weight":106,"slug":136},19,"General","general"]