{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/spatiotemporal-feature-integration-and-model","title":"SpatioTemporal Feature Integration and Model Fusion for Full Reference Video Quality Assessment","arxiv_id":"1804.04813","date":"2018-04-13","proceeding":null,"authors":[],"abstract":"Perceptual video quality assessment models are either frame-based or\nvideo-based, i.e., they apply spatiotemporal filtering or motion estimation to\ncapture temporal video distortions. Despite their good performance on video\nquality databases, video-based approaches are time-consuming and harder to\nefficiently deploy. To balance between high performance and computational\nefficiency, Netflix developed the Video Multi-method Assessment Fusion (VMAF)\nframework, which integrates multiple quality-aware features to predict video\nquality. Nevertheless, this fusion framework does not fully exploit temporal\nvideo quality measurements which are relevant to temporal video distortions. To\nthis end, we propose two improvements to the VMAF framework: SpatioTemporal\nVMAF and Ensemble VMAF. Both algorithms exploit efficient temporal video\nfeatures which are fed into a single or multiple regression models. To train\nour models, we designed a large subjective database and evaluated the proposed\nmodels against state-of-the-art approaches. The compared algorithms will be\nmade available as part of the open source package in\nhttps://github.com/Netflix/vmaf.","url_abs":"http://arxiv.org/abs/1804.04813v1","url_pdf":"http://arxiv.org/pdf/1804.04813v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"spatiotemporal-feature-integration-and-model","repo_url":"https://github.com/Netflix/vmaf","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"computational-efficiency","task_name":"Computational Efficiency"},{"task_slug":"motion-estimation","task_name":"Motion Estimation"},{"task_slug":"video-quality-assessment","task_name":"Video Quality Assessment"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}