[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84967-en":3,"doc-seo-84967-105":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},84967,7971461740886,"Theodore","https://ap-avatar.wpscdn.com/davatar_3d24733baf745e90a7e4bdd5f77d97b2",8,"Research & Report","TACoS Weakly Supervised Learning of Two-Dimensional Materials from Scribble Annotations to Precise Segmentation","Precise pixel-level localization of two-dimensional material flakes is essential for high-throughput screening, yet dense annotations required by fully supervised segmentation are costly and slow, hindering real deployment. TACoS introduces a tree-based asymmetric contrast segmentation framework for scribble supervision. It combines unlabeled weak–strong consistency learning with tree energy regularization to derive structure-aware soft pseudo labels. Asymmetric regional contrast learning merges scribbles with high-confidence weak predictions and uses category prototypes to improve boundary stability.","arXiv :2607 .07 169v 1 [ cs .CV] 8 Jul 2026  \nTACoS: Weakly Supervised Learning of Two-Dimensional Materials from Scribble Annotations to Precise Segmentation  \nJiabei Chena,c , Liping Zhanga,c , Jiang-Bin Wub,c , Zhongming Weib,c , Enhao Ninga , Su Yana,c , Weijun Lia,c , Ping-Heng Tanb,c , Xin Ninga,c,∗  \na AnnLab, Institute of Semiconductors, Chinese Academy of Sciences, Beijing, 100083, China b State Key Laboratory of Semiconductor Physics and Chip Technologies, Institute of Semiconductors,  \nChinese Academy of Sciences, Beijing, 100083, China c Center of Materials Science and Optoelectronics Engineering & School of Integrated Circuits, University of Chinese Academy of Sciences, Beijing, 100049, China  \nAbstract  \nThe precise pixel-level localization of two-dimensional material flakes is crucial for high-throughput screening. However, traditional fully supervised methods rely on dense annotations, which are costly and time-consuming, severely limiting the practical deployment of segmentation models. This paper proposes Tree-based Asymmetric Contrast Segmentation (TACoS), a specialized scribble segmentation framework tailored for two-dimensional materials. First, we design a unified framework that integrates semi-supervised consistency learning with structured tree energy constraints. This framework comprises two core components: an unlabeled weak-strong distribution alignment module and a tree energy regularization module. The former employs cosine consistency constraints to enhance prediction alignment across views. Meanwhile, the latter utilizes minimum spanning trees to establish pixel affinity relationships and generate structure-aware soft pseudo labels for online semantic guidance. Next, we introduce asymmetric regional contrast learning. This approach fuses high-confidence predictions from the weak augmentation branch with scribbles to form augmented labels, and construct category prototypes in the representation space. Simultaneously, we prioritize contrastive constraints on challenging pixels in boundary-unlabeled re-  \n∗ Corresponding author  \nEmail address: [ningxin@semi.ac.cn](ningxin@semi.ac.cn) (Xin Ning)  \ngions. This strategy enhances intra-class cohesion and inter-class separation at the representation level, effectively reducing category confusion in low-contrast edges and complex backgrounds. Experiments conducted on the constructed graphene and MoS2 datasets demonstrate that our method TACoS achieves over 96% of fully supervised performance using less than 0.6% annotated data. Furthermore, it exhibits superior structural coherence and boundary stability in scenarios with weakly contrasting edgesand complex backgrounds, providing an efficient and scalable solution for automated high-throughput screening of two-dimensional material flakes.  \nKeywords: two-dimensional materials, optical microscopic, sparse annotation, weakly supervised learning, semantic segmentation  \n1. Introduction  \nSince graphene was successfully isolated [1], two-dimensional materials have rapidly become a major research focus in condensed-matter and materials physics. Experimentally, mechanical exfoliation [2] is widely used to prepare two-dimensional material samples due to its simplicity and high crystalline quality. Taking advantage of thinfilm interference [3], exfoliated flakes exhibit obvious reflectance contrast and clear morphological cues under visible light, which provides a physically observable basis for pixel-level flake segmentation from microscopic images.  \nIn recent years, semantic segmentation has been successfully applied to automated identification of two-dimensional material flakes [4, 5] . When training images are sufficiently annotated, models can learn optical contrast, edge gradients, and texture patterns of target regions, enabling precise segmentation at pixel-level. However, in real laboratory settings, images are often large, have complex backgrounds, and contain diverse impurities. Pixel-level annotatio","cbCaikEbICkZJy5p","https://ap.wps.com/l/cbCaikEbICkZJy5p","pdf",1091175,1,35,"English","en",105,"# Introduction\n## Motivation and background\n## Scribble-based weak supervision\n## Related work and limitations","[{\"question\":\"What problem does TACoS address in two-dimensional material segmentation?\",\"answer\":\"TACoS targets the high cost and inefficiency of dense pixel-level annotations required by fully supervised segmentation for two-dimensional material microscopic images.\"},{\"question\":\"How does TACoS leverage scribble annotations?\",\"answer\":\"TACoS uses an asymmetric framework where high-confidence predictions from a weak augmentation branch are fused with scribbles to create augmented labels, guiding semantic segmentation with limited supervision.\"},{\"question\":\"What performance and scalability results are reported?\",\"answer\":\"Experiments on graphene and MoS2 datasets show TACoS reaches over 96% of fully supervised performance while using less than 0.6% annotated data, and improves structural coherence and boundary stability under complex backgrounds.\"}]",1784199805,88,{"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},"tacos-weakly-supervised-learning-of-two-dimensional-materials-from-scribble-annotations-to-precise-segmentation","",{"@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/tacos-weakly-supervised-learning-of-two-dimensional-materials-from-scribble-annotations-to-precise-segmentation/84967/",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 TACoS address in two-dimensional material segmentation?","Question",{"text":75,"@type":76},"TACoS targets the high cost and inefficiency of dense pixel-level annotations required by fully supervised segmentation for two-dimensional material microscopic images.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does TACoS leverage scribble annotations?",{"text":80,"@type":76},"TACoS uses an asymmetric framework where high-confidence predictions from a weak augmentation branch are fused with scribbles to create augmented labels, guiding semantic segmentation with limited supervision.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance and scalability results are reported?",{"text":84,"@type":76},"Experiments on graphene and MoS2 datasets show TACoS reaches over 96% of fully supervised performance while using less than 0.6% annotated data, and improves structural coherence and boundary stability under complex backgrounds.","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"]