[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86115-en":3,"doc-seo-86115-105":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},86115,1374391974468,"Eden","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",6,"Technology","Realtime Face Video Enhancement (RTFVE)","RTFVE (Realtime Face Video Enhancement) addresses the quality limits of video conferencing under low bandwidth by introducing a face-enhancement model designed for practical deployment. The paper targets two barriers in existing deep learning approaches: difficulty integrating with common video compression pipelines and high compute demands requiring specialized GPUs or NPUs. RTFVE can be incorporated into a video decoder and run in realtime on ordinary CPUs, improving perceptual face quality across multiple low-bitrate settings while keeping compute feasible.","arXiv :2607 . 11034v1 [ cs .CV] 13 Jul 2026  \nRTFVE: Realtime Face Video Enhancement  \nVarun Ramesh Jois, Antonella DiLillo, James Storer  \nBrandeis University, Waltham MA 02453, USA  \n{vjois,dilant,[storer}@brandeis.edu](storer}@brandeis.edu)  \nAbstract. There’s been a surge in adoption of video conferencing applications for both personal and business use cases. However, the bandwidth limitations faced by many users worldwide may restrict the optimal use of such applications. Although deep learning offers a solution for enhancing low bit rate videos, most models today are either hard to incorporate with modern compression standards or require specialized hardware to run such as significant GPUs making these models impractical. To address these issues, we introduce the Realtime Face Video Enhancement (RTFVE) model which can be easily incorporated with any video decoderand can run in realtime on ordinary CPUs. Experiments show that our model improves perceptual quality over the compressed video baseline at multiple low bitrate settings. The source code will be made available at [https://github.com/varun-jois/RTFVE](https://github.com/varun-jois/RTFVE).  \nKeywords: Face Enhancement · Realtime · Videoconferencing.  \n1 Introduction  \nOver the last decade videocalling has become one of the most popular forms of communication. With today’s technology, everyone has access to a phone with a front-facing camera and an internet connection, often making video calling a daily occurrence for many users. Ever since the lockdowns during the COVID-19 pandemic, the adoption of this technology has grown exponentially with many new companies providing services for it.  \nWhile there has been a global trend towards faster internet speeds, there are still millions of people around the world that experience slow internet connections. And while the video compression standards are flexible with handling low bandwidth settings, they come with the cost of low quality video quite often making videocalls untenable. This being the case even though many of these users have the basic hardware to handle a standard videocall.  \nDeep learning based methods hold promise when it comes to improving the perceptual quality of faces. There have been advances in a myriad of fields such as face restoration, face deblurring, face super-resolution, etc. However, when it comes to face enhancement for videocalls there have been two problems stymieing the adoption of deep learning: 1) Models that cannot be easily integrated into the ubiquitous compression standards and 2) Models that require significant GPU and NPU resources that typical users don’t have. In this paper, we address both of these issues.  \nThis is a post-peer-review, pre-copyedit version of an article published in CAIP 2025 . The final authenticated version is  \navailable online at [https://doi.org/10.1007/978-3-032-04968-1_12](https://doi.org/10.1007/978-3-032-04968-1_12) .  \n2 Jois et al.  \nTable 1 . Relative speed comparison of various methods ranging from classical computer vision algorithms to current state-of-the-art face restoration models on a low cost CPU. For realtime performance an FPS of at least 24 is required.  \n\n| Model | Type | Frames per Second (↑) |\n| --- | --- | --- |\n| Non-Local Means Denoising [3] | Classical | 4.56 |\n| Bilateral Filtering [17] | Classical | 13.89 |\n| Split-Bregman [6] | Classical | 17.71 |\n| Chambolle [4] | Classical | 7.60 |\n| Codeformer [23] | Neural Network | 0.05 |\n| GPEN [21] | Neural Network | 2.47 |\n| GFP-GAN [19] | Neural Network | 2.02 |\n| RTFVE (Ours) | Neural Network | 24.85 |\n\nOur contributions are as follows:  \n1. We present our model RealTime Face Video Enhance-ment (RTFVE) a model for face enhancement that can easily be integrated with the decoder of a video compression standard.  \n2. For the bandwidth cost of a handful of high-quality reference frames, our model is able to improve the visual quality of faces over the compression standards running in realt","cbCaimxzgHLyjmWX","https://ap.wps.com/l/cbCaimxzgHLyjmWX","pdf",2213550,1,12,"English","en",105,"# Introduction\n## Face enhancement for videocalls\n# Related Work\n## Face restoration\n## Video restoration","[{\"question\":\"What problem does RTFVE address in video conferencing?\",\"answer\":\"RTFVE targets the degraded face quality caused by low bitrate and bandwidth limitations in video calls. It focuses on improving perceptual quality of faces under compressed video conditions.\"},{\"question\":\"Why are existing deep learning face enhancement models hard to adopt for videocalls?\",\"answer\":\"The paper identifies two issues: models that cannot be easily integrated with widely used compression standards, and models that require significant GPU/NPU resources that typical users do not have.\"},{\"question\":\"How does RTFVE achieve realtime performance for most users?\",\"answer\":\"RTFVE is designed to be integrated with a video decoder and to run in realtime on ordinary CPUs. Experimental results report realtime FPS performance, with perceptual improvements over the compressed baseline at multiple low bitrate settings.\"}]",1784208614,30,{"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},"realtime-face-video-enhancement-rtfve","",{"@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/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/realtime-face-video-enhancement-rtfve/86115/",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","2026-07-17","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 RTFVE address in video conferencing?","Question",{"text":75,"@type":76},"RTFVE targets the degraded face quality caused by low bitrate and bandwidth limitations in video calls. It focuses on improving perceptual quality of faces under compressed video conditions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why are existing deep learning face enhancement models hard to adopt for videocalls?",{"text":80,"@type":76},"The paper identifies two issues: models that cannot be easily integrated with widely used compression standards, and models that require significant GPU/NPU resources that typical users do not have.",{"name":82,"@type":73,"acceptedAnswer":83},"How does RTFVE achieve realtime performance for most users?",{"text":84,"@type":76},"RTFVE is designed to be integrated with a video decoder and to run in realtime on ordinary CPUs. Experimental results report realtime FPS performance, with perceptual improvements over the compressed baseline at multiple low bitrate settings.","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},[93,97,101,105,110,113,118,122,127,130,134],{"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":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":114,"doc_module":4,"doc_module_name":45,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":45,"category_name":120,"show_sort_weight":28,"slug":121},8,"Research & Report","research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]