{"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/high-order-attention-models-for-visual","title":"High-Order Attention Models for Visual Question Answering","arxiv_id":"1711.04323","date":"2017-11-12","proceeding":"NeurIPS 2017 12","authors":["Idan Schwartz","Alexander G. Schwing","Tamir Hazan"],"abstract":"The quest for algorithms that enable cognitive abilities is an important part\nof machine learning. A common trait in many recently investigated\ncognitive-like tasks is that they take into account different data modalities,\nsuch as visual and textual input. In this paper we propose a novel and\ngenerally applicable form of attention mechanism that learns high-order\ncorrelations between various data modalities. We show that high-order\ncorrelations effectively direct the appropriate attention to the relevant\nelements in the different data modalities that are required to solve the joint\ntask. We demonstrate the effectiveness of our high-order attention mechanism on\nthe task of visual question answering (VQA), where we achieve state-of-the-art\nperformance on the standard VQA dataset.","url_abs":"http://arxiv.org/abs/1711.04323v1","url_pdf":"http://arxiv.org/pdf/1711.04323v1.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":"high-order-attention-models-for-visual","repo_url":"https://github.com/idansc/HighOrderAtten","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"torch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"},{"task_slug":"high","task_name":"Vocal Bursts Intensity Prediction"}],"methods":[{"method_slug":"fga","method_name":"FGA"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-question-answering-on-coco-visual-1","task":"Visual Question Answering (VQA)","dataset":"COCO Visual Question Answering (VQA) real images 1.0 multiple choice","model":"3-Modalities: Unary + Pairwise + Ternary (ResNet)","rank_in_archive_order":4,"of":10,"metrics":{"Percentage correct":"69.3"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1711.04323","atlas_url":"https://app.syntology.ai/?focus=1711.04323","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}