[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85231-en":3,"doc-seo-85231-105":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},85231,13056703019662,"Evangeline","https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188",8,"Research & Report","Anomalous Frame Detection by Grouping Frame Similarities between Two Videos Computed by Vision-Language Model","Maintenance of critical infrastructures like railways and power plants demands reliable operation, yet the number of skilled maintenance workers keeps declining, making efficient transfer of expert know-how urgent. Traditional interview methods cannot capture tacit aspects experts may not consciously articulate. This work proposes anomalous frame detection by comparing manual-based work videos with videos from expert maintenance workers. In a simulated distribution board task, 11 action types missing from the manual were extracted with a 66.9% success rate, outperforming conventional methods by 50 percentage points.","Anomalous Frame Detection by Grouping Frame Similarities between Two Videos Computed by Vision-Language Model to Extract Expert Workers’Unique Actions  \nRyo Sakai1*, Yongpeng Cao2* and Nobutaka Kimura3  \n1 Robotics Research Department, Research and Development Group, Hitachi, Ltd., Ibaraki, Japan  \n2 School of Engineering, The University of Tokyo, Tokyo, Japan  \n3 Next Research Department, Research and Development Group, Hitachi, Ltd., Kokubunji, Japan Corresponding author: Ryo Sakai (e-mail: [ryo.sakai.cx@hitachi.com](ryo.sakai.cx@hitachi.com)).  \nABSTRACT Maintenance of critical infrastructures, such as railways and power plants, is essential for operational safety and reliability. However, the declining number of skilled maintenance workers poses a serious challenge to sustaining these operations, highlighting the need to effectively transfer expert knowhow to less experienced workers. Although traditional interview-based approaches have been used to elicit maintenance skills, they struggle to capture know-how that experts themselves may not consciously recognize. To address this gap, we proposed a method that detects anomalous frames of candidate actions including know-how by comparing a video of manual-based work with that of expert maintenance workers. In a simulated maintenance experiment involving a distribution board, our method targeted 11 types of actions not described in the manual and achieved a 66.9% extraction rate, marking a 50-percentage-point improvement over conventional techniques. These findings underscore the effectiveness of our approach in revealing hidden maintenance knowledge, thereby contributing to enhanced skill transfer and workforce development in critical infrastructure maintenance.  \nINDEX TERMS Anomalous frame detection, Similarities between videos, Expert worker ’s action, VisionLanguage Model  \nI. INTRODUCTION  \nMaintenance inspection work is one of the critical tasks in the operation of infrastructure facilities such as railwaysand power plants. Since infrastructure facilities undergo repeated repairs and renovations over their long service lives, the maintenance tasks themselves have become increasingly complex. Meanwhile, there is a shortage of skilled maintenance personnel capable of performing these tasks adequately at each site [1, 2, 3]. One potential solution to this shortage is the training of new maintenance inspectors. Although new inspectors require time to reexperience the work performed by expert inspectors in order to acquire equivalent maintenance know-how, it is expected that understanding the expert’s operational knowhow could significantly shorten this period—as seen, for example, in cases where expert know-how in ship  \n*Equal contribution.  \noperations is used to accelerate personnel training [4] . In general, however, such expert know-how is often tacit and held only by the individual expert. If this know-how can be codified, it would be possible to educate maintenance procedures systematically, thereby efficiently increasing the pool of potential maintenance inspectors [5] . From these perspectives, there is a growing need to extract the work know-how from experts, convert it into explicit  \nknowledge, and pass it on to less experienced workers.  \nOne approach to codifying the work know-how possessed by experts is the interview technique employed by previous study [4, 6, 7, 8, 9, 10, 11] . In this approach, an interviewer prepares work-related questions and extracts the expert’s know-how from their responses. This technique can only codify the work know-how that the interviewer can anticipate or that the expert is consciously  \naware of. On the other hand, it is challenging to obtain information regarding aspects of know-how that the interviewer does not anticipate or that the expert is unaware of. Moreover, actions that have become second nature to the expert—even if they are critical to the quality of work—may not be deliberately explained during the interview and th","cbCaisEayAKgyJPg","https://ap.wps.com/l/cbCaisEayAKgyJPg","pdf",1174681,1,11,"English","en",105,"# Introduction\n## Motivation and need for know-how extraction\n## Limits of interview-based approaches\n## Video-based extraction and related differences detection\n## Goal and proposed anomalous frame detection approach","[{\"question\":\"What problem does the study address?\",\"answer\":\"The study addresses the declining availability of skilled maintenance workers and the resulting difficulty in transferring expert maintenance know-how to less experienced workers.\"},{\"question\":\"How does the proposed method detect expert know-how?\",\"answer\":\"It compares videos of manual-based work with videos of expert operations, then detects anomalous frames within candidate actions that reflect differences from the manual.\"},{\"question\":\"What were the experimental results and how were they evaluated?\",\"answer\":\"In a simulated maintenance experiment involving a distribution board, the method targeted 11 action types not described in the manual and achieved a 66.9% extraction rate, a 50-percentage-point improvement over conventional techniques.\"}]",1784201886,28,{"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},"anomalous-frame-detection-by-grouping-frame-similarities-between-two-videos-computed-by-vision-language-model","",{"@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/anomalous-frame-detection-by-grouping-frame-similarities-between-two-videos-computed-by-vision-language-model/85231/",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","2026-07-24","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 the study address?","Question",{"text":75,"@type":76},"The study addresses the declining availability of skilled maintenance workers and the resulting difficulty in transferring expert maintenance know-how to less experienced workers.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the proposed method detect expert know-how?",{"text":80,"@type":76},"It compares videos of manual-based work with videos of expert operations, then detects anomalous frames within candidate actions that reflect differences from the manual.",{"name":82,"@type":73,"acceptedAnswer":83},"What were the experimental results and how were they evaluated?",{"text":84,"@type":76},"In a simulated maintenance experiment involving a distribution board, the method targeted 11 action types not described in the manual and achieved a 66.9% extraction rate, a 50-percentage-point improvement over conventional techniques.","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"]