[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-141417-105":59,"doc-detail-141417-en":134},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":127,"head_meta":129,"extra_data":131,"updated_unix":133},105,"en","predictive-quality-assessment-for-mobile-secure-graphics-iccv-workshop-2025-paper","Predictive Quality Assessment for Mobile Secure Graphics - ICCV Workshop 2025 paper","","Secure graphic verification, a key anti-counterfeiting mechanism, suffers from unreliable smartphone image acquisition that drives high false rejection rates and creates a “reliability gap.” The work replaces traditional perceptual image quality assessment with proactive predictive quality estimation: a lightweight model scores video frames by their expected utility for downstream verification. The approach is validated on 32,000+ images from 105 smartphones using re-contextualized FNMR and ISRR metrics. Cross-domain experiments show a frozen, ImageNet-pretrained backbone generalizes better to unseen printing technologies than full fine-tuning, improving robustness under domain shift.",{"@graph":69,"@context":126},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/predictive-quality-assessment-for-mobile-secure-graphics-iccv-workshop-2025-paper/141417/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/predictive-quality-assessment-for-mobile-secure-graphics-iccv-workshop-2025-paper/141417.png","ImageObject",300,407,{"name":92,"@type":93},"Evangeline","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-09-21","2026-08-25",true,{"@type":102,"interactionType":103,"userInteractionCount":34},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118,122],{"name":109,"@type":110,"acceptedAnswer":111},"What reliability problem does the paper focus on for secure graphic verification?","Question",{"text":112,"@type":113},"It targets the reliability gap caused by uncontrolled smartphone captures, where poor lighting, motion blur, or defocus increase false rejections and can even make genuine graphics statistically indistinguishable from counterfeits.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does the proposed method differ from traditional no-reference image quality assessment?",{"text":117,"@type":113},"Instead of predicting perceptual quality from texture statistics independent of verification, it learns a lightweight model that predicts a frame’s utility for the downstream verification task.",{"name":119,"@type":110,"acceptedAnswer":120},"How is the method evaluated and what metrics are used?",{"text":121,"@type":113},"It evaluates the module on a large-scale dataset (32,000+ images from 105 smartphones) using re-contextualized FNMR and ISRR, enabling a system-level trade-off analysis between reliability and usability.",{"name":123,"@type":110,"acceptedAnswer":124},"What does the cross-domain analysis reveal about model generalization?",{"text":125,"@type":113},"A lightweight probe on a frozen, ImageNet-pretrained network generalizes better to unseen printing technologies than a fully fine-tuned model, suggesting frozen general-purpose backbones are more robust under manufacturing domain shifts.","https://schema.org",{"og:url":83,"og:type":128,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":130,"canonical":83},"index,follow",{"doc_id":132,"site_id":62},141417,1787655723,{"code":4,"msg":5,"data":135},{"doc_id":132,"user_id":136,"nickname":92,"user_avatar":137,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":138,"file_id":139,"file_url":140,"file_type":141,"file_size":142,"view_count":34,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":52,"language":143,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":144,"faqs":145,"seo_title":146,"seo_description":67,"update_tm":133,"read_time":147},13056703019662,"https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188","This ICCV Workshop paper is the Open Access version, provided by the Computer Vision Foundation.  \nExcept for this watermark, it is identical to the accepted version; the final published version of the proceedings is available on IEEE Xplore.  \nPredictive Quality Assessment for Mobile Secure Graphics  \nCas Steigstra  \nScantrust / University of Amsterdam [cas.steigstra@gmail.com](cas.steigstra@gmail.com)  \nSergey Milyaev Scantrust  \n[sergey.milyaev@gmail.com](sergey.milyaev@gmail.com)  \nShaodi You University of Amsterdam  \n[s.you@uva.nl](s.you@uva.nl)  \nMobile Capture: Image Sequence  \n0 2 4 6 8 10  \nFrame  \nFigure 1 . Our Predictive Score Aligns with a Secure Graphic’s Quality. A mobile capture session progresses from noisy (left) to high-fidelity (right) . Our predictive score (Ours, blue) correctly tracks this improvement. In contrast, standard IQA methods (dashed lines) prefer the initial noisy frames, giving a misleading signal about the frame’s utility for verification. This shows the need for a task-specific approach over general-purpose IQA.  \nAbstract  \nThe reliability of secure graphic verification, a key anticounterfeiting tool, is undermined by poor image acquisition on smartphones. Uncontrolled user captures of these high-entropy patterns cause high false rejection rates, creating a significant ‘reliability gap’. To bridge this gap, we depart from traditional perceptual IQA and introduce a framework that predictively estimates a frame’s utility for the downstream verification task. We propose a lightweight model to predict a quality score for a video frame, determining its suitability for a resource-intensive oracle model. Our framework is validated using re-contextualized FNMR and ISRR metrics on a large-scale dataset of 32,000+ images from 105 smartphones. Furthermore, a novel crossdomain analysis on graphics from different industrial printing presses reveals a key finding: a lightweight probe on a  \nfrozen, ImageNet-pretrained network generalizes better to an unseen printing technology than a fully fine-tuned model. This provides a key insight for real-world generalization:  \nfor domain shifts from physical manufacturing, a frozen general-purpose backbone can be more robust than full finetuning, which can overfit to source-domain artifacts.  \n1. Introduction  \nCounterfeit goods pose a formidable global challenge, inflicting trillions in economic damage and threatening public safety [9] . Physical-digital verification systems, which leverage the ubiquity of smartphones to verify physical products, have emerged as a promising solution. At the forefront of this are camera-verifiable security graphics. As illustrated in Figure 2, our work focuses on a common implementation of this technology: a high-entropy secure graphic embedded within a standard QR code [24, 35] . This hybrid design creates a secure, unclonable link to a digital identity that can be printed directly onto products.  \nCurrent research in security graphic verification has primarily focused on the security of the patterns themselves and the development of classifiers to distinguish genuine  \nitems from sophisticated fakes [2, 5, 14] . These works, while crucial, almost universally operate under a strong, implicit assumption: that the captured image is of sufficient quality for reliable analysis. This task is highly sensitive to image quality, creating a “minimum quality gap” in realworld use. Our work addresses this foundational quality problem. Users capture images under uncontrolled conditions—poor lighting, motion blur, and defocus—leading to high false rejection rates [23, 32] . More critically, severe image degradation can erode the statistical features of a genuine graphic to the point where it becomes indistinguishable from a counterfeit, turning a usability issue into a fundamental security vulnerability. Resolving this gap is crucial for the widespread adoption of such visual verification systems. Our work addresses this challenge within the","cbCaifVdJpbTcdlp","https://ap.wps.com/l/cbCaifVdJpbTcdlp","pdf",4472615,"English","# Abstract\n## Predictive quality estimation framework\n## Evaluation setup and dataset\n## Cross-domain analysis and findings\n# Introduction\n## Problem: minimum quality gap and false rejections\n## Proposed proactive two-stage paradigm\n## Trade-off via FNMR and ISRR","[{\"question\":\"What reliability problem does the paper focus on for secure graphic verification?\",\"answer\":\"It targets the reliability gap caused by uncontrolled smartphone captures, where poor lighting, motion blur, or defocus increase false rejections and can even make genuine graphics statistically indistinguishable from counterfeits.\"},{\"question\":\"How does the proposed method differ from traditional no-reference image quality assessment?\",\"answer\":\"Instead of predicting perceptual quality from texture statistics independent of verification, it learns a lightweight model that predicts a frame’s utility for the downstream verification task.\"},{\"question\":\"How is the method evaluated and what metrics are used?\",\"answer\":\"It evaluates the module on a large-scale dataset (32,000+ images from 105 smartphones) using re-contextualized FNMR and ISRR, enabling a system-level trade-off analysis between reliability and usability.\"},{\"question\":\"What does the cross-domain analysis reveal about model generalization?\",\"answer\":\"A lightweight probe on a frozen, ImageNet-pretrained network generalizes better to unseen printing technologies than a fully fine-tuned model, suggesting frozen general-purpose backbones are more robust under manufacturing domain shifts.\"}]","Predictive Quality Assessment for Mobile Secure Graphics - ICCV Workshop 2025 paper | PDF",25]