[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85469-en":3,"doc-seo-85469-105":29,"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":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},85469,549758146520,"Patrick","https://ap-avatar.wpscdn.com/avatar/80002397d8c0411e94?_k=1775819394049821470",8,"Research & Report","Towards Blind Lens Aberration Correction via Large LensLib Pre training and Discrete Degradation Priors","LensLib-PT pre-training enables blind lens aberration correction by learning a universal network for diverse unknown optical degradations. The paper introduces FoundCAC to overcome generalization limits caused by insufficient data scalability and missing prior guidance for optical degradation. It builds AODLibpro using stratified sampling over spatial-variation patterns and degradation severity, and proposes a multi-stage vector-quantized scheme to form a Latent PSF Representation, discretizing continuous PSFs into a prior that regularizes restoration. Experiments show state-of-the-art zero-shot results and efficient few-shot adaptation.","arXiv :2511 . 17126v5 [ ee ss .IV] 11 Jul 2026  \nTowards Blind Lens Aberration Correction via Large LensLib Pre-training and Discrete  \nDegradation Priors  \nXiaolong Qian∗ , Qi Jiang∗ , Yao Gao, Lei SunB , Kailun Yang, Xian Wang, Zhonghua Yi, Wenyong Li,  \nMing-Hsuan Yang, Luc Van Gool, and Kaiwei WangB  \nAbstract—Emerging deep-learning-based lens library pre-training (LensLib-PT) pipeline offers a new avenue for blind lens aberration correction by training a universal neural network, demonstrating strong capability in handling diverse unknown optical degradations. This work proposes FoundCAC, a universal foundational framework that resolves two challenges hindering the generalization of existing pipelines: the difficulty of scaling training data and the absence of prior guidance characterizing optical degradation. To improve data scalability, we expand the design specifications to increase degradation diversity and construct AODLibpro, a large-scale lens library using stratified sampling over spatial-variation patterns and degradation severity. In terms of model design, to leverage Point Spread Functions (PSFs) as guidance while maintaining the blind paradigm, we propose a multi-stage vector-quantized representation learning scheme. This paradigm is specifically designed to construct a Latent PSF Representation (LPR), explicitly encoding complex continuous PSFs into a discrete degradation prior to regularize the highly ill-posed restoration process. Through a simple yet effective codebookfreezing strategy, our framework leverages the discrete prior to elevate full-shot restoration performance and unlock highly efficient few-shot adaptation for unseen lenses. Experiments on synthetic LensLib, real-design simulations, and real-captured lenses show that our framework achieves state-of-the-art zero-shot performance under complementary evaluation protocols, while enabling highly efficient few-shot adaptation for specific lenses. The source code and datasets will be made publicly available at FoundCAC.  \nIndex Terms—Computational Photography, Lens Aberration Correction, Point Spread Function, Vector-Quantized Representation  \n~~ ~~ ✦ ~~ ~~  \n1 INTRODUCTION  \nLENS aberrations, typically arising from compromised  \nimage quality optimization due to design trade-offs for specific requirements, e.g., minimalist optical systems [2],[3], or lenses on mobile devices [4], and manufacturing/assembly errors [5] in complex systems, introduce blur to the captured images. This blur is also referred to as optical degradation [6], characterized by its distinctive spatiallyvarying nature where degradation varies across Field-ofViews (FoVs) and exhibits diverse patterns depending on optical path, representing a fundamental image quality issue but has received limited attention in the learning and vision literature. With the advancement of image processing, computational post-processing [7] has become a mainstream pipeline, also known as computational aberration correction. Unlike non-blind methods that rely on precise Point Spread Functions (PSFs) calibration [8], the blind pipeline [9] offers more flexible and user-friendly advantages for users without optical expertise, where only the captured images are required for high-quality results.  \n• X. Qian, Q. Jiang, Y. Gao, L. Sun, X. Wang, Z. Yi, W. Li, and K. Wang are with the National Research Center for Optical Instrumentation, Zhejiang University, Hangzhou 310027, China.  \n• L. Sun is also with INSAIT, Sofia University “St. Kliment Ohridski”, Sofia 1784, Bulgaria.  \n• K. Yang is with the School of Artificial Intelligence and Robotics, Hunan University, Changsha 410012, China.  \n• K. Yang is also with the National Engineering Research Center of Robot Visual Perception and Control Technology, Hunan University, Changsha 410082, China.  \n• ∗Equal contribution.  \n• B Correspondence (E-mail: leo [sun@zju.edu.cn](sun@zju.edu.cn), [wangkaiwei@zju.edu.cn](wangkaiwei@zju.edu.cn)).  \nRecently, the deep learn","cbCaifqYRhhAsTOg","https://ap.wps.com/l/cbCaifqYRhhAsTOg","pdf",19310292,1,29,"English","en",105,"# Introduction\n## Lens aberrations and optical degradation\n## Lens Library Pre-Training pipelines\n## Proposed FoundCAC framework\n## Data construction and AODLibpro\n## Model design with Latent PSF Representation","[{\"question\":\"What problem does FoundCAC address in blind lens aberration correction?\",\"answer\":\"FoundCAC targets two bottlenecks that limit LensLib-PT generalization: biased and poorly scalable training data, and lack of explicit prior guidance describing optical degradation for end-to-end restoration.\"},{\"question\":\"How does AODLibpro improve data scalability and realism?\",\"answer\":\"AODLibpro increases degradation diversity by expanding design specifications and uses stratified sampling over spatial-variation patterns and degradation severity to better capture real-world aberration behavior.\"},{\"question\":\"How is point spread function (PSF) information used while keeping the pipeline blind?\",\"answer\":\"The method leverages PSFs as guidance through a multi-stage vector-quantized representation learning approach that constructs a Latent PSF Representation, discretizing continuous PSFs into a degradation prior to regularize the ill-posed restoration.\"}]",1784203804,73,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":27},"towards-blind-lens-aberration-correction-via-large-lenslib-pre-training-and-discrete-degradation-priors","",{"@graph":35,"@context":85},[36,53,68],{"@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/towards-blind-lens-aberration-correction-via-large-lenslib-pre-training-and-discrete-degradation-priors/85469/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","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 FoundCAC address in blind lens aberration correction?","Question",{"text":75,"@type":76},"FoundCAC targets two bottlenecks that limit LensLib-PT generalization: biased and poorly scalable training data, and lack of explicit prior guidance describing optical degradation for end-to-end restoration.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does AODLibpro improve data scalability and realism?",{"text":80,"@type":76},"AODLibpro increases degradation diversity by expanding design specifications and uses stratified sampling over spatial-variation patterns and degradation severity to better capture real-world aberration behavior.",{"name":82,"@type":73,"acceptedAnswer":83},"How is point spread function (PSF) information used while keeping the pipeline blind?",{"text":84,"@type":76},"The method leverages PSFs as guidance through a multi-stage vector-quantized representation learning approach that constructs a Latent PSF Representation, discretizing continuous PSFs into a degradation prior to regularize the ill-posed restoration.","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":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]