[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81679-en":3,"doc-seo-81679-105":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},81679,1649267921044,"Ava Thompson","https://us-avatar.wpscdn.com/avatar/1800007509477c92dfb?_k=1782875107921204101",8,"Research & Report","CUPID Reconstructing UV Texture Maps for Interpretable Person-of-Interest Deepfake Detection","Deepfakes targeting a specific Person-of-Interest (POI) threaten modern societies, yet existing POI detectors often trade off robustness to post-processing, computational efficiency, and interpretability. This paper introduces CUPID, a POI video deepfake detector combining UV texture maps from 3D face reconstructions with Masked Autoencoder (MAE) representation learning. Training requires no POI videos. At inference, embeddings from a query POI video are matched to pristine references, and UV-space residual maps localize identity deviations. Experiments on four datasets show state-of-the-art performance, strong robustness to downscaling and compression, and faster inference.","CUPID: Reconstructing UV Texture Maps for Interpretable Person-of-Interest Deepfake Detection  \nGiovanni Affatato Student Member, IEEE, Sara Mandelli Member, IEEE, Edoardo Daniele Cannas Member, IEEE, Paolo Bestagini Member, IEEE, and Stefano Tubaro Senior Member, IEEE  \narXiv :2606 .20302v2 [ cs .CV] 9 Jul 2026  \nAbstract—Deepfakes targeting a high-profile individual, known as Person-of-Interest (POI), are a threat to modern democracies and societies. Current POI deepfake detection methods still struggle to combine robustness to post-processing, efficiency and interpretability, key aspects of modern deepfake detectors. In this paper we propose CUPID, a POI video deepfake detector that combines UV texture maps, a facial appearance representation derived from 3D face reconstructions, with the representation learning capabilities of the Masked Autoencoder (MAE).  \nOur method does not require any deepfake videos in its training phase. Moreover, it does not even require including a specific POI in the training set: the combination of UV texture maps extracted from real video frames and the MAE contextguided reconstruction yields a latent space that captures rich and discriminative facial features even for identities unseen during training. In the testing phase, the embeddings extracted from a query video depicting the POI can be matched against pristine reference videos to assess the video authenticity. Furthermore, operating in the UV space naturally provides an additional layer of interpretability. Specifically, we can extract decoded residual maps that highlight which facial regions of a test video deviate most from the identity representation of the corresponding POI.  \nExperiments on four deepfake datasets show that CUPID outperforms the current state of the art on most datasets and achieves the best overall robustness against strong downscaling and compression, while also providing substantially faster inference. Our experimental code will be released at polimiispl/CUPID.  \nIndex Terms—person-of-interest deepfake detection, masked autoencoder, 3D morphable models, interpretability, robustness.  \nI. INTRODUCTION  \nIn recent years, the rapid advancements in the field of generative AI (GenAI) have enabled the creation of synthetic videos, popularly known as deepfakes, of unprecedented realism [1] . The malicious use of GenAI technologies has fueled the spread of deepfakes targeting specific individuals, such as politicians [2], celebrities [3], and high-profile corporate employees [4], with the aim of spreading misinformation and enabling fraudulent activities.  \nIn response, the forensics community has developed specialized methods to protect a specific individual, referred to asthe Person-of-Interest (POI), by exploiting the large amount  \nThe authors are with the Dipartimento di Elettronica, Informazione e Bioingegneria (DEIB), Politecnico di Milano, 20133 Milan, Italy. This work was supported by the FOSTERER project, funded by the Italian Ministry of Education, University, and Research within the PRIN 2022 program. This work was partially supported by the European Union -Next Generation EU under the Italian National Recovery and Resilience Plan (NRRP), Mission 4, Component 2, Investment 1.3, CUP D43C22003080001, partnership on“Telecommunications of the Future” (PE00000001 - program “RESTART”) and by the Investment 1.3, CUP D43C22003050001, partnership on “SEcurity and RIghts in the CyberSpace”(PE00000014-program “FF4ALL-SERICS”) .  \nFig. 1. Overview of the proposed CUPID method for POI deepfake detection. During training (top), a masked autoencoder encodes UV texture maps from real videos of many subjects into a general latent representation of facial identity. During inference (bottom), the same encoder maps pristine videos of the POI into this latent space to build a reference identity distribution. A test video, whether real or manipulated, is then encoded and compared against this distribution for authenticity assessmen","cbCaiqSSeZNperXP","https://ap.wps.com/l/cbCaiqSSeZNperXP","pdf",4349173,2,1,16,"English","en",105,"# Introduction\n## POI deepfake detection background and challenges\n## Related approaches: training-time POI vs inference-time POI\n## Proposed method: CUPID pipeline idea","[{\"question\":\"What problem does CUPID address in POI deepfake detection?\",\"answer\":\"CUPID targets the difficulty of building POI detectors that remain robust to post-processing while also achieving efficiency and interpretability.\"},{\"question\":\"How does CUPID handle training without using POI-specific videos?\",\"answer\":\"CUPID does not require any deepfake videos during training and does not need the specific POI included in the training set; it uses UV texture maps from real frames and MAE context-guided reconstruction to form a discriminative latent space.\"},{\"question\":\"How is interpretability achieved during testing?\",\"answer\":\"CUPID operates in the UV space and can extract decoded residual maps to highlight which facial regions deviate most from the POI identity representation.\"}]",1784175374,40,{"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},"cupid-reconstructing-uv-texture-maps-for-interpretable-person-of-interest-deepfake-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/cupid-reconstructing-uv-texture-maps-for-interpretable-person-of-interest-deepfake-detection/81679/",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","2026-07-24","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 problem does CUPID address in POI deepfake detection?","Question",{"text":75,"@type":76},"CUPID targets the difficulty of building POI detectors that remain robust to post-processing while also achieving efficiency and interpretability.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does CUPID handle training without using POI-specific videos?",{"text":80,"@type":76},"CUPID does not require any deepfake videos during training and does not need the specific POI included in the training set; it uses UV texture maps from real frames and MAE context-guided reconstruction to form a discriminative latent space.",{"name":82,"@type":73,"acceptedAnswer":83},"How is interpretability achieved during testing?",{"text":84,"@type":76},"CUPID operates in the UV space and can extract decoded residual maps to highlight which facial regions deviate most from the POI identity representation.","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},[93,97,101,105,110,115,119,122,127,130,134],{"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":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":29,"slug":118},7,"Healthcare","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":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]