{"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/structure-propagation-for-zero-shot-learning","title":"Structure propagation for zero-shot learning","arxiv_id":"1711.09513","date":"2017-11-27","proceeding":null,"authors":["Guangfeng Lin","Yajun Chen","Fan Zhao"],"abstract":"The key of zero-shot learning (ZSL) is how to find the information transfer\nmodel for bridging the gap between images and semantic information (texts or\nattributes). Existing ZSL methods usually construct the compatibility function\nbetween images and class labels with the consideration of the relevance on the\nsemantic classes (the manifold structure of semantic classes). However, the\nrelationship of image classes (the manifold structure of image classes) is also\nvery important for the compatibility model construction. It is difficult to\ncapture the relationship among image classes due to unseen classes, so that the\nmanifold structure of image classes often is ignored in ZSL. To complement each\nother between the manifold structure of image classes and that of semantic\nclasses information, we propose structure propagation (SP) for improving the\nperformance of ZSL for classification. SP can jointly consider the manifold\nstructure of image classes and that of semantic classes for approximating to\nthe intrinsic structure of object classes. Moreover, the SP can describe the\nconstrain condition between the compatibility function and these manifold\nstructures for balancing the influence of the structure propagation iteration.\nThe SP solution provides not only unseen class labels but also the relationship\nof two manifold structures that encode the positive transfer in structure\npropagation. Experimental results demonstrate that SP can attain the promising\nresults on the AwA, CUB, Dogs and SUN databases.","url_abs":"http://arxiv.org/abs/1711.09513v1","url_pdf":"http://arxiv.org/pdf/1711.09513v1.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":"structure-propagation-for-zero-shot-learning","repo_url":"https://github.com/lgf78103/Structure-propagation-for-zero-shot-learning","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"zero-shot-learning","task_name":"Zero-Shot Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}