[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-293391-105":59,"doc-detail-293391-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","mbti-masked-blending-transformers-with-implicit-positional-encoding-for-frame-rate-agnostic-motion-estimation","MBTI - Masked Blending Transformers with Implicit Positional Encoding for Frame-rate Agnostic Motion Estimation","","Human motion estimation models often assume a fixed number of input frames, making them sensitive to changes in frame rate and producing inconsistent motion predictions across different temporal resolutions. This issue stems from frame rates defining temporal granularity, so models trained for one sampling frequency struggle under another. MBTI (Masked Blending Transformers with Implicit Positional Encoding) aligns inputs to a high reference frame rate using masked token blending and introduces implicit positional encoding for absolute time-aware reconstruction. It also evaluates using EMDB-FPS and the MCF metric, showing improved accuracy and temporal consistency across frame rates.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/mbti-masked-blending-transformers-with-implicit-positional-encoding-for-frame-rate-agnostic-motion-estimation/293391/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/mbti-masked-blending-transformers-with-implicit-positional-encoding-for-frame-rate-agnostic-motion-estimation/293391.png","ImageObject",300,407,{"name":92,"@type":93},"Stanford","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-29","2026-09-17",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",16,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"Why do motion estimation models become inconsistent when the input frame rate changes?","Question",{"text":113,"@type":114},"Most models assume a fixed number of input frames, so the actual temporal coverage varies with the frame rate. This leads to temporal distortions and misaligned motion interpretation across different sampling frequencies.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"What key components does MBTI introduce to handle varying frame rates?",{"text":118,"@type":114},"MBTI uses a masked autoencoder with masked token blending to align tokens to a high reference frame rate, and adds implicit positional encoding to encode absolute time via neural implicit functions.",{"name":120,"@type":111,"acceptedAnswer":121},"How are MBTI’s robustness and temporal consistency evaluated?",{"text":122,"@type":114},"The paper introduces EMDB-FPS, an augmented benchmark for multiple frame rates in local and global tasks, and proposes MCF, a metric quantifying deviation in motion predictions across different input frame rates.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},293391,1789663490,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":130,"read_time":145,"preview_page_count":4},2336477552062,"https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d","This ICCV paper is the Open Access version, provided by the Computer Vision Foundation.  \nExcept for this watermark, it is identical to the accepted version; the final published version of the proceedings is available on IEEE Xplore.  \nMBTI: Masked Blending Transformers with Implicit Positional Encoding for  \nFrame-rate Agnostic Motion Estimation  \nJungwoo Huh Yeseung Park Seongjean Kim Jungsu Kim Sanghoon Lee∗  \nYonsei University, Seoul, South Korea  \n{gjwjddn9,pys940617,jin.k,integer,[slee](slee}@yonsei.ac.kr)[}](slee}@yonsei.ac.kr)[@yonsei.ac.kr](slee}@yonsei.ac.kr)  \nFigure 1. Frame-rate Agnostic Motion Estimation. State-of-the-art motion estimation models (WHAM [41], GVHMR [39], TRAM [47]) fail to estimate consistent motion from the same video across different frame rates. The results of each frame rate are aligned at the same absolute time positions. While WHAM fails to estimate incosistent camera trajectory, GVHMR and TRAM fails to align poses across different frame rate input. In contrast, MBTI successfully estimates consistent motion regardless of the input video frame rate.  \nAbstract  \nHuman motion estimation models typically assume a fixed number of input frames, making them sensitive to variations in frame rate and leading to inconsistent motion predictions across different temporal resolutions. This limitation arises because input frame rates inherently determine the temporal granularity of motion capture, causing discrepancies when models trained on a specific frame rate encounter different sampling frequencies. To address this challenge, we propose MBTI (Masked Blending Transformers with Implicit Positional Encoding), a frame rate-agnostic human motion estimation framework designed to maintain temporal consistency across varying input frame rates. Our approach leverages a masked autoencoder (MAE) architecture with masked token blending, which aligns input tokens with a predefined high-reference frame rate, ensuring a  \n*Corresponding author.  \nstandardized temporal representation. Additionally, we introduce implicit positional encoding, which encodes absolute time information using neural implicit functions, enabling more natural motion reconstruction beyond discrete sequence indexing. By reconstructing motion at a high reference frame rate and optional downsampling, MBTI ensures both frame rate generalization and temporal consistency. To comprehensively evaluate MBTI, we introduce EMDB-FPS, an augmented benchmark designed to assess motion estimation robustness across multiple frame rates in both local and global motion estimation tasks. To further assess MBTI’s robustness, we introduce the Motion Consistency across Frame rates (MCF), a novel metric to quantify the deviation of motion predictions across different input frame rates. Our results demonstrate that MBTI outperforms state-of-the-art methods in both motion accuracy and temporal consistency, achieving the most stable and consistent motion predictions across varying frame rates.  \n1. Introduction  \nHuman motion estimation from video remains an active research area, spanning from local motion estimation in camera coordinates [13, 14, 21, 23, 35] to global motion estimation in world coordinates [19, 20, 25, 27, 41, 47, 52, 55], where both human movement and camera motion are considered. In most cases, these methods rely on video frames as their primary input. However, videos are inherently discretely sampled signals, with the frame rate determining their temporal resolution. The choice of frame rate varies by application: low-frame rate videos are commonly used in bandwidth-limited environments or surveillance systems [10, 22], whereas moderate-frame rate videos are typical for general-purpose cameras. In contrast, high-frame rate videos (e.g., 120 FPS or more) are frequently used in motion capture systems for fine-grained motion analysis [31, 36] . Although real-world human motion is continuous, human motion estimation models must reconstruct natural mot","cbCaiksiWyDV1w6N","https://ap.wps.com/l/cbCaiksiWyDV1w6N","pdf",3426558,11,"English","# Abstract\n# 1. Introduction","[{\"question\":\"Why do motion estimation models become inconsistent when the input frame rate changes?\",\"answer\":\"Most models assume a fixed number of input frames, so the actual temporal coverage varies with the frame rate. This leads to temporal distortions and misaligned motion interpretation across different sampling frequencies.\"},{\"question\":\"What key components does MBTI introduce to handle varying frame rates?\",\"answer\":\"MBTI uses a masked autoencoder with masked token blending to align tokens to a high reference frame rate, and adds implicit positional encoding to encode absolute time via neural implicit functions.\"},{\"question\":\"How are MBTI’s robustness and temporal consistency evaluated?\",\"answer\":\"The paper introduces EMDB-FPS, an augmented benchmark for multiple frame rates in local and global tasks, and proposes MCF, a metric quantifying deviation in motion predictions across different input frame rates.\"}]","MBTI - Masked Blending Transformers with Implicit Positional Encoding for Frame-rate Agnostic Motion Estimation | PDF",28]