[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85273-en":3,"doc-seo-85273-105":29,"detail-sidebar-cat-0-en-105":82},{"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":4,"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},85273,1374391974585,"Genevieve","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Temporal Feature Distillation for Label Efficient Precise Event Spotting in Sports Videos","Precise Event Spotting (PES) distinguishes visually similar but semantically different adjacent frames, requiring boundary-level temporal localization that differs from image classification and coarse action recognition. Directly transferring self-distillation methods like DINO to PES suppresses subtle yet crucial motion cues, producing representations insensitive to event transitions. Temporal Feature Distillation (TFD) uses a semisupervised objective to align temporally informative backbone features for localization. A supervised warm-up with ramp-up stabilizes training before unlabeled distillation, and Transformer Gate Shift injects multi-scale motion-aware information. Experiments on four benchmarks show consistent mAP gains; with under 10% supervision on FSPerf, mAP improves by 4.54.","Temporal Feature Distillation for Label-Efficient Precise Event  \nSpotting in Sports Videos  \nHao Xu  \nDeakin University Melbourne, Australia [august.xu@research.deakin.edu.au](august.xu@research.deakin.edu.au)  \nXinyu Wei  \nChampion Data Melbourne, Australia [felix.wei@championdata.com.au](felix.wei@championdata.com.au)  \nSam Wells  \nParalympics Australia Melbourne, Australia [sam.wells@paralympic.org.au](sam.wells@paralympic.org.au)  \nSunil Aryal  \nDeakin University Melbourne, Australia [sunil.aryal@deakin.edu.au](sunil.aryal@deakin.edu.au)  \narXiv :2607 . 10998v1 [ cs .CV] 13 Jul 2026  \nAbstract  \nPrecise Event Spotting (PES) requires distinguishing visually similar yet semantically distinct adjacent frames, making it fundamentally different from image classification and coarse action recognition. Although self-distillation methods such as DINO have shown strong representation learning ability in images, we find that directly applying them to PES is ineffective: without supervised guidance, subtle but crucial motion cues are often suppressed as noise, leading to representations that are insensitive to precise event boundaries. To address this, we propose Temporal Feature Distillation, a semisupervised objective that aligns temporally informative backbone features, rather than projection-head outputs, to preserve motionsensitive and boundary-aware cues for frame-level localization. A supervised warm-up with a ramp-up schedule further stabilizestraining by ensuring that meaningful event cues are learned before unlabeled distillation begins. We also introduce Transformer Gate Shift, a multi-scale gated shifting module that injects motionaware temporal information into Vision Transformers. Experiments on four fine-grained sports benchmarks show consistent improvements over fully supervised and semi-supervised baselines. Under 10% supervision on FSPerf, our method improves mAP by 4.54 points over the strongest competing approach, and with only 80% labeled data, it matches or surpasses the fully supervised 100% baseline on two of the four datasets. Code is available at the anonymous repository: [https://anonymous.4open.science/r/TFD-8535](https://anonymous.4open.science/r/TFD-8535) .  \nCCS Concepts  \n• Computing methodologies → Computer vision.  \nKeywords  \nPrecise event spotting, semi-supervised learning, sports video understanding  \n1 Introduction  \nWith the rapid growth of broadcast and amateur recordings, sports videos have become an important source for performance analysis, coaching, and audience engagement [36] . Among sports video understanding tasks, detecting the precise moment of key events plays a fundamental role in interpreting game dynamics [6] .  \nPrecise Event Spotting (PES) is a task specifically designed for sports videos, aiming to identify both the temporal location and the  \ncategory of events—such as a forehand stroke in tennis or a pass in soccer—from long, untrimmed video sequences [6] . Unlike coarse action recognition, PES requires frame-level temporal precision and faces unique challenges inherent to sports scenarios, including rapid motion, frequent occlusion, camera motion, and visually subtle events [6, 30, 31] .  \nAlthough recent fully supervised methods have achieved promising results on multiple sports datasets [6, 16, 26, 31, 37], their reliance on dense frame-level annotations limits scalability [17] . Annotating sports videos is time-consuming, often requires expert knowledge, and can be inconsistent when event boundaries are ambiguous. The challenge is even greater in amateur sports videos, where lower frame rates make precise timestamp annotation more difficult.  \nAnother defining characteristic of sports videos is their continuous and large-scale generation, as new matches are recorded daily across professional and amateur levels. Effectively leveraging this ever-growing pool of unlabeled data is therefore essential. In parallel, Unsupervised (USL) and semi-supervised learning (SSL) have ga","cbCaidIKtRgw4pxn","https://ap.wps.com/l/cbCaidIKtRgw4pxn","pdf",1157480,1,9,"English","en",105,"# Introduction\n## Precise Event Spotting in sports videos\n## Limitations of fully supervised and existing self-distillation\n## Proposed Temporal Feature Distillation (TFD)\n# Method and contributions","[{\"question\":\"How effective is the method under limited supervision in experiments?\",\"answer\":\"On FSPerf, under 10% supervision the method increases mAP by 4.54 points over the strongest competing approach. With only 80% labeled data, it matches or surpasses the fully supervised 100% baseline on two of the four datasets.\"}]",1784202199,23,{"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":77,"head_meta":79,"extra_data":81,"updated_unix":27},"temporal-feature-distillation-for-label-efficient-precise-event-spotting-in-sports-videos","",{"@graph":35,"@context":76},[36,53,67],{"@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/temporal-feature-distillation-for-label-efficient-precise-event-spotting-in-sports-videos/85273/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70],{"name":71,"@type":72,"acceptedAnswer":73},"How effective is the method under limited supervision in experiments?","Question",{"text":74,"@type":75},"On FSPerf, under 10% supervision the method increases mAP by 4.54 points over the strongest competing approach. With only 80% labeled data, it matches or surpasses the fully supervised 100% baseline on two of the four datasets.","Answer","https://schema.org",{"og:url":51,"og:type":78,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":80,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":83},[84,88,92,96,101,106,111,114,118,121,125],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":45,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":45,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":112,"slug":113},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":115,"show_sort_weight":116,"slug":117},"Religion & Spirituality",20,"religion-spirituality",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":116,"slug":120},"World Cup","world-cup",{"id":122,"doc_module":4,"doc_module_name":45,"category_name":123,"show_sort_weight":122,"slug":124},10,"Lifestyle","lifestyle",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":97,"slug":128},19,"General","general"]