[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-138302-en":3,"doc-seo-138302-105":30,"detail-sidebar-cat-0-en-105":92},{"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":27,"seo_description":14,"update_tm":28,"read_time":29},138302,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",6,"Technology","Sign Language Recognition using Sub-Units","This paper discusses sign language recognition using linguistic sub-units, including sub-units learned from appearance data and those inferred from 2D or 3D tracking. Three sub-unit types are evaluated and combined into sign-level classifiers using two main strategies: Markov Models that encode temporal changes between sub-units, and Sequential Pattern Boosting that performs discriminative feature selection while capturing temporal information. The proposed approach is more robust to noise and achieves stronger signer-independent results than Markov Chains.","Sign Language Recognition using Sub-Units  \nHelen Cooper H. M. COOPER@ SURREY. AC . UK  \nEng-Jon Ong E. ONG@ SURREY. AC . UK  \nNicolas Pugeault N. PUGEAULT@ SURREY. AC . UK  \nRichard Bowden R. BOWDEN@ SURREY. AC . UK  \nCentre for Vision Speech and Signal Processing University of Surrey  \nGuildford. GU2 9PY UK  \nEditors: Isabelle Guyon and Vassilis Athitsos  \nAbstract  \nThis paper discusses sign language recognition using linguistic sub-units. It presents three types of sub-units for consideration; those learnt from appearance data as well as those inferred from both 2D or 3D tracking data. These sub-units are then combined using a sign level classi􀀂er; here, two options are presented. The 􀀂rst uses Markov Models to encode the temporal changes between sub-units. The second makes use of Sequential Pattern Boosting to apply discriminative featureselection at the same time as encoding temporal information. This approach is more robust to noise and performs well in signer independent tests, improving results from the 54% achieved by the Markov Chains to 76% .  \nKeywords: sign language recognition, sequential pattern boosting, depth cameras, sub-units, signer independence, data set  \n1. Introduction  \nThis paper presents several approaches to sub-unit based Sign Language Recognition (SLR) culminating in a real time KinectTM demonstration system. SLR is a non-trivial task. Sign Languages (SLs) are made up of thousands of different signs; each differing from the other by minor changes in motion, handshape, location or Non-Manual Featuress (NMFs) . While Gesture Recognition (GR) solutions often build a classi􀀂er per gesture, this approach soon becomes intractable when recognising large lexicons of signs, for even the relatively straightforward task of citation-form, dictionary look-up. Speech recognition was faced with the same problem; the emergent solution was to recognise the subcomponents (phonemes), then combine them into words using Hidden Markov Models (HMMs) . Sub-unit based SLR uses a similar two stage recognition system, in the 􀀂rst stage, sign linguistic sub-units are identi􀀂ed. In the second stage, these sub-units are combined together to create a sign level classi􀀂er.  \nLinguists also describe SLs in terms of component sub-units; by using these sub-units, not only can larger sign lexicons be handled ef􀀂ciently, allowing demonstration on databases of nearly 1000 signs, but they are also more robust to the natural variations of signs, which occur on both an inter and an intra signer basis. This makes them suited to real-time signer independent recognition as described later. This paper will focus on 4 main sub-unit categories based on HandShape, Location, Motion and Hand-Arrangement. There are several methods for labelling these sub-units and this  \n􀀍c2012 Helen Cooper, Nicolas Pugeault, Eng-Jon Ong and Richard Bowden.  \nCOOPER, PUGEAULT, ONG AND BOWDEN  \nFigure 1: Overview of the 3 types of sub-units extracted and the 2 different sign level classi􀀂ers used.  \nwork builds on both the Ha, Tab, Sig, Dez system from the BSL dictionary (British Deaf Association, 1992) and The Hamburg Notation System (HamNoSys), which has continued to develop over recent years to allow more detailed description of signs from numerous SLs (Hanke and Schmaling, 2004) .  \nThis paper presents a comparison of sub-unit approaches, focussing on the advantages and disadvantages of each. Also presented is a newly released Kinect data set, containing multiple users performing signs in various environments. There are three different types of sub-units considered; those based on appearance data alone, those which use 2D tracking data with appearance based handshapes and those which use 3D tracking data produced by a KinectTM sensor. Each of these  \nSIGN LANGUAGE RECOGNITION USING SUB-UNITS  \nthree sub-unit types is tested with a Markov model approach to combine sub-units into sign level classi􀀂ers. A further experiment is performed to investigate the discrim","cbCaiabidCNR7uLj","https://ap.wps.com/l/cbCaiabidCNR7uLj","pdf",2388079,1,27,"English","en",105,"# Introduction\n## Sub-unit based sign language recognition\n# Background\n## Prior work and sub-unit modeling approaches\n# Methodology\n## Markov model combination of sub-units\n## Sequential Pattern Boosting for signer-independent recognition","[{\"question\":\"What are the three types of sub-units considered in the paper?\",\"answer\":\"The paper considers sub-units based on appearance data alone, sub-units using 2D tracking with appearance-based handshapes, and sub-units using 3D tracking produced by a Kinect sensor.\"},{\"question\":\"How are sub-units combined into sign-level classifiers?\",\"answer\":\"Sub-units are combined using sign-level classifiers through two options: Markov Models to encode temporal transitions, or Sequential Pattern Boosting to jointly encode temporal information and perform discriminative feature selection.\"},{\"question\":\"What performance improvement is reported for signer-independent recognition?\",\"answer\":\"The method reports signer-independent results of 76%, improving over 54% achieved by Markov Chains.\"}]","Sign Language Recognition using Sub-Units | PDF",1787478734,68,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"sign-language-recognition-using-sub-units","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/technology/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/sign-language-recognition-using-sub-units/138302/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-31","2026-08-23",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What are the three types of sub-units considered in the paper?","Question",{"text":76,"@type":77},"The paper considers sub-units based on appearance data alone, sub-units using 2D tracking with appearance-based handshapes, and sub-units using 3D tracking produced by a Kinect sensor.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How are sub-units combined into sign-level classifiers?",{"text":81,"@type":77},"Sub-units are combined using sign-level classifiers through two options: Markov Models to encode temporal transitions, or Sequential Pattern Boosting to jointly encode temporal information and perform discriminative feature selection.",{"name":83,"@type":74,"acceptedAnswer":84},"What performance improvement is reported for signer-independent recognition?",{"text":85,"@type":77},"The method reports signer-independent results of 76%, improving over 54% achieved by Markov Chains.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,114,119,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":112,"slug":113},50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":120,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":122,"slug":123},8,"Research & Report",30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]