[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-1-en-105":3,"doc-seo-238206-105":53,"doc-detail-238206-en":126},{"code":4,"msg":5,"data":6},0,"success",[7,14,19,24,29,34,39,44,49],{"id":8,"doc_module":9,"doc_module_name":10,"category_name":11,"show_sort_weight":12,"slug":13},11,1,"Template","Presentations",90,"presentations",{"id":15,"doc_module":9,"doc_module_name":10,"category_name":16,"show_sort_weight":17,"slug":18},12,"Resumes",80,"resumes",{"id":20,"doc_module":9,"doc_module_name":10,"category_name":21,"show_sort_weight":22,"slug":23},14,"Invoices",70,"invoices",{"id":25,"doc_module":9,"doc_module_name":10,"category_name":26,"show_sort_weight":27,"slug":28},15,"Posters",60,"posters",{"id":30,"doc_module":9,"doc_module_name":10,"category_name":31,"show_sort_weight":32,"slug":33},16,"Social Media",50,"social-media",{"id":35,"doc_module":9,"doc_module_name":10,"category_name":36,"show_sort_weight":37,"slug":38},17,"Forms",40,"forms",{"id":40,"doc_module":9,"doc_module_name":10,"category_name":41,"show_sort_weight":42,"slug":43},18,"Letters",30,"letters",{"id":45,"doc_module":9,"doc_module_name":10,"category_name":46,"show_sort_weight":47,"slug":48},21,"Paper Templates",5,"papers-templates",{"id":50,"doc_module":9,"doc_module_name":10,"category_name":51,"show_sort_weight":4,"slug":52},158,"General","general-158",{"code":4,"msg":54,"data":55},"ok",{"site_id":56,"language":57,"slug":58,"title":59,"keywords":60,"description":61,"schema_data":62,"social_meta":119,"head_meta":121,"extra_data":123,"updated_unix":125},105,"en","autotransition-learning-to-recommend-video-transition-effects","AutoTransition - Learning to Recommend Video Transition Effects","","Video transition effects are widely used to connect shots and create cohesive videos, yet selecting suitable transitions remains difficult for non-professionals lacking cinematography knowledge and design skills. This paper introduces automatic video transition recommendation (VTR), which takes a sequence of raw video shots and companion audio and recommends transitions for each neighboring shot pair. A large-scale transition dataset is collected from public editing templates. VTR is formulated as a multi-modal retrieval problem using a multi-modal transformer for vision/audio fusion and sequential context modeling, validated by quantitative and qualitative experiments and a user study.",{"@graph":63,"@context":118},[64,80,101],{"@type":65,"itemListElement":66},"BreadcrumbList",[67,71,74,77],{"item":68,"name":69,"@type":70,"position":9},"https://docshare.wps.com","Home","ListItem",{"item":72,"name":10,"@type":70,"position":73},"https://docshare.wps.com/template/",2,{"item":75,"name":11,"@type":70,"position":76},"https://docshare.wps.com/template/presentations/",3,{"item":78,"name":59,"@type":70,"position":79},"https://docshare.wps.com/template/autotransition-learning-to-recommend-video-transition-effects/238206/",4,{"url":78,"name":59,"@type":81,"image":82,"author":87,"headline":59,"publisher":90,"fileFormat":93,"inLanguage":57,"description":61,"dateModified":94,"datePublished":95,"encodingFormat":93,"isAccessibleForFree":96,"interactionStatistic":97},"DigitalDocument",{"url":83,"@type":84,"width":85,"height":86},"https://docshare.wps.com/thumbnails/autotransition-learning-to-recommend-video-transition-effects/238206.png","ImageObject",442,249,{"name":88,"@type":89},"Ophelia","Person",{"url":68,"name":91,"@type":92},"DocShare","Organization","application/pdf","2026-10-22","2026-09-11",true,{"@type":98,"interactionType":99,"userInteractionCount":25},"InteractionCounter",{"@type":100},"ViewAction",{"@type":102,"mainEntity":103},"FAQPage",[104,110,114],{"name":105,"@type":106,"acceptedAnswer":107},"What problem does AutoTransition address?","Question",{"text":108,"@type":109},"AutoTransition addresses automatic video transition recommendation (VTR), recommending transitions for each pair of neighboring video shots given raw shots and companion audio.","Answer",{"name":111,"@type":106,"acceptedAnswer":112},"How is VTR defined compared with conventional classification?",{"text":113,"@type":109},"Unlike single-label classification, VTR produces a ranked list of candidate transition categories so users can choose freely during video editing.",{"name":115,"@type":106,"acceptedAnswer":116},"How does the approach learn to match vision/audio inputs to transitions?",{"text":117,"@type":109},"The method formulates VTR as multi-modal retrieval: it learns transition embeddings via a video transition classification task, then uses a multi-modal transformer to fuse vision and audio and model sequential context.","https://schema.org",{"og:url":78,"og:type":120,"og:title":59,"og:site_name":91,"og:description":61},"article",{"robots":122,"canonical":78},"index,follow",{"doc_id":124,"site_id":56},238206,1791748189,{"code":4,"msg":5,"data":127},{"doc_id":124,"user_id":128,"nickname":88,"user_avatar":129,"doc_module":9,"category_id":8,"category_name":11,"doc_title":59,"doc_description":61,"doc_content":130,"file_id":131,"file_url":132,"file_type":133,"file_size":134,"view_count":25,"is_deleted":4,"is_public":9,"is_downloadable":9,"audit_status":9,"page_count":135,"language":136,"language_code":57,"site_id":56,"html_lang":57,"table_of_contents":137,"faqs":138,"seo_title":139,"seo_description":61,"update_tm":140,"read_time":141},7971461741311,"https://ap-avatar.wpscdn.com/avatar/74000253aff267980c6?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345379180704826","arXiv :2207 . 13479v1 [ cs .CV] 27 Jul 2022  \nAutoTransition: Learning to Recommend Video  \nTransition E􀀋ects  \nYaojie Shen 1 ;2 ;3 􀀃 , Libo Zhang 1 ;2 􀀃 , Kai Xu3 􀀃 , Xiaojie Jin3􀀃y  \n1 Institute of Software, Chinese Academy of Sciences  \n2 University of Chinese Academy of Sciences  \n3 ByteDance Inc.  \nAbstract. Video transition e􀀋ects are widely used in video editing to connect shots for creating cohesive and visually appealing videos. However, it is challenging for non-professionals to choose best transitions due to the lack of cinematographic knowledge and design skills. In this paper, we present the premier work on performing automatic video transitions recommendation (VTR): given a sequence of raw video shots and companion audio, recommend video transitions for each pair of neighboring shots. To solve this task, we collect a large-scale video transition dataset using publicly available video templates on editing softwares. Then we formulate VTR as a multi-modal retrieval problem from vision/audio to video transitions and propose a novel multi-modal matching framework which consists of two parts. First we learn the embedding of video transitions through a video transition classi􀀌cation task. Then we propose a model to learn the matching correspondence from vision/audio inputs to video transitions. Speci􀀌cally, the proposed model employs a multi-modal transformer to fuse vision and audio information, as well as capture the context cues in sequential transition outputs. Through both quantitative and qualitative experiments, we clearly demonstrate the e􀀋ectiveness of our method. Notably, in the comprehensive user study, our method receives comparable scores compared with professional editors while improving the video editing e􀀎ciency by 300􀀂 . We hope our work serves to inspire other researchers to work on this new task. The dataset and codes are public at [https://github.com/acherstyx/AutoTransition](https://github.com/acherstyx/AutoTransition).  \nKeywords: video transition e􀀋ects recommendation, multi-modal retrieval, video editing  \n1 Introduction  \nWith the advance of multimedia technology and network infrastructures, video is ubiquitous, occurring in numerous everyday activities such as education, entertainment, surveillance, etc. There is a massive amount of needs for people to edit videos and share with others. However, video editing is challenging for nonprofessionals since it is not only laborious but also needs a lot of cinematography  \n*  \nEqual contribution.  \n† Project lead and corresponding author \u003C[jinxiaojie@bytedance.com](jinxiaojie@bytedance.com) > .  \n2 Y. Shen et al.  \nand design knowledge. Some editing tools like Adobe Premier and Apple Final Cut Pro are developed to assist video editing, however their main target users are professionals while novices may 􀀌nd it di􀀎cult to learn. Moreover, they still lack the ability of automatic video editing, i.e., users have to manipulate videoson their own. Recently, popular video editing tools like InShot Video Editor and CapCut provide the function of creating videos in one-click. Nevertheless, since they only utilize simple strategies or 􀀌xed video templates and ignore the content of input vision/audio, the quality of generated video is unsatisfactory.  \nThe video transition e􀀋ects play an important role in video editing to join shots for creating smooth and cohesive videos. In this paper, we introduce anew task of automatic video transition recommendation (VTR) and provide a systematic solution. Speci􀀌cally, VTR is de􀀌ned as: given a sequence of raw video shots and the companion audio (which can either be original sound or overwritten music), recommend a sequence of video transitions for each neighboring shots. Di􀀋erent from conventional classi􀀌cation problems which choose only one most probable category as the output, VTR aims to provide a ranking of candidate transition categories so that users can choose freely in practical usage.  \nWhen working on this task, ","cbCaiuEAiHLFSIHj","https://ap.wps.com/l/cbCaiuEAiHLFSIHj","pdf",9082448,22,"English","# Introduction\n## Problem Definition\n## Challenges\n## Dataset Construction\n## Method Overview and Experiments","[{\"question\":\"What problem does AutoTransition address?\",\"answer\":\"AutoTransition addresses automatic video transition recommendation (VTR), recommending transitions for each pair of neighboring video shots given raw shots and companion audio.\"},{\"question\":\"How is VTR defined compared with conventional classification?\",\"answer\":\"Unlike single-label classification, VTR produces a ranked list of candidate transition categories so users can choose freely during video editing.\"},{\"question\":\"How does the approach learn to match vision/audio inputs to transitions?\",\"answer\":\"The method formulates VTR as multi-modal retrieval: it learns transition embeddings via a video transition classification task, then uses a multi-modal transformer to fuse vision and audio and model sequential context.\"}]","AutoTransition - Learning to Recommend Video Transition Effects | PDF",1789132265,8]