{"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/factorizable-net-an-efficient-subgraph-based","title":"Factorizable Net: An Efficient Subgraph-based Framework for Scene Graph Generation","arxiv_id":"1806.11538","date":"2018-06-29","proceeding":"ECCV 2018 9","authors":["Yikang Li","Wanli Ouyang","Bolei Zhou","Jianping Shi","Chao Zhang","Xiaogang Wang"],"abstract":"Generating scene graph to describe all the relations inside an image gains\nincreasing interests these years. However, most of the previous methods use\ncomplicated structures with slow inference speed or rely on the external data,\nwhich limits the usage of the model in real-life scenarios. To improve the\nefficiency of scene graph generation, we propose a subgraph-based connection\ngraph to concisely represent the scene graph during the inference. A bottom-up\nclustering method is first used to factorize the entire scene graph into\nsubgraphs, where each subgraph contains several objects and a subset of their\nrelationships. By replacing the numerous relationship representations of the\nscene graph with fewer subgraph and object features, the computation in the\nintermediate stage is significantly reduced. In addition, spatial information\nis maintained by the subgraph features, which is leveraged by our proposed\nSpatial-weighted Message Passing~(SMP) structure and Spatial-sensitive Relation\nInference~(SRI) module to facilitate the relationship recognition. On the\nrecent Visual Relationship Detection and Visual Genome datasets, our method\noutperforms the state-of-the-art method in both accuracy and speed.","url_abs":"http://arxiv.org/abs/1806.11538v2","url_pdf":"http://arxiv.org/pdf/1806.11538v2.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":"factorizable-net-an-efficient-subgraph-based","repo_url":"https://github.com/yikang-li/FactorizableNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"graph-generation","task_name":"Graph Generation"},{"task_slug":"relationship-detection","task_name":"Relationship Detection"},{"task_slug":"scene-graph-generation","task_name":"Scene Graph Generation"},{"task_slug":"visual-relationship-detection","task_name":"Visual Relationship Detection"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/scene-graph-generation-on-vrd","task":"Scene Graph Generation","dataset":"VRD","model":"FactorizableNet","rank_in_archive_order":1,"of":2,"metrics":{"Recall@50":"18.32"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.11538","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}