{"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/semantic-video-cnns-through-representation","title":"Semantic Video CNNs through Representation Warping","arxiv_id":"1708.03088","date":"2017-08-10","proceeding":"ICCV 2017 10","authors":["Raghudeep Gadde","Varun Jampani","Peter V. Gehler"],"abstract":"In this work, we propose a technique to convert CNN models for semantic\nsegmentation of static images into CNNs for video data. We describe a warping\nmethod that can be used to augment existing architectures with very little\nextra computational cost. This module is called NetWarp and we demonstrate its\nuse for a range of network architectures. The main design principle is to use\noptical flow of adjacent frames for warping internal network representations\nacross time. A key insight of this work is that fast optical flow methods can\nbe combined with many different CNN architectures for improved performance and\nend-to-end training. Experiments validate that the proposed approach incurs\nonly little extra computational cost, while improving performance, when video\nstreams are available. We achieve new state-of-the-art results on the CamVid\nand Cityscapes benchmark datasets and show consistent improvements over\ndifferent baseline networks. Our code and models will be available at\nhttp://segmentation.is.tue.mpg.de","url_abs":"http://arxiv.org/abs/1708.03088v1","url_pdf":"http://arxiv.org/pdf/1708.03088v1.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":"semantic-video-cnns-through-representation","repo_url":"https://github.com/raghudeep/netwarp_public","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"optical-flow-estimation","task_name":"Optical Flow Estimation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.03088","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}