{"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/a-spatiotemporal-oriented-energy-network-for","title":"A Spatiotemporal Oriented Energy Network for Dynamic Texture Recognition","arxiv_id":"1708.06690","date":"2017-08-22","proceeding":"ICCV 2017 10","authors":["Isma Hadji","Richard P. Wildes"],"abstract":"This paper presents a novel hierarchical spatiotemporal orientation\nrepresentation for spacetime image analysis. It is designed to combine the\nbenefits of the multilayer architecture of ConvNets and a more controlled\napproach to spacetime analysis. A distinguishing aspect of the approach is that\nunlike most contemporary convolutional networks no learning is involved;\nrather, all design decisions are specified analytically with theoretical\nmotivations. This approach makes it possible to understand what information is\nbeing extracted at each stage and layer of processing as well as to minimize\nheuristic choices in design. Another key aspect of the network is its recurrent\nnature, whereby the output of each layer of processing feeds back to the input.\nTo keep the network size manageable across layers, a novel cross-channel\nfeature pooling is proposed. The multilayer architecture that results\nsystematically reveals hierarchical image structure in terms of multiscale,\nmultiorientation properties of visual spacetime. To illustrate its utility, the\nnetwork has been applied to the task of dynamic texture recognition. Empirical\nevaluation on multiple standard datasets shows that it sets a new\nstate-of-the-art.","url_abs":"http://arxiv.org/abs/1708.06690v1","url_pdf":"http://arxiv.org/pdf/1708.06690v1.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":"a-spatiotemporal-oriented-energy-network-for","repo_url":"https://github.com/hadjisma/soe-net","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"dynamic-texture-recognition","task_name":"Dynamic Texture Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}