{"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/image-question-answer-synergistic-network-for","title":"Image-Question-Answer Synergistic Network for Visual Dialog","arxiv_id":"1902.09774","date":"2019-02-26","proceeding":"CVPR 2019 6","authors":["Dalu Guo","Chang Xu","DaCheng Tao"],"abstract":"The image, question (combined with the history for de-referencing), and the\ncorresponding answer are three vital components of visual dialog. Classical\nvisual dialog systems integrate the image, question, and history to search for\nor generate the best matched answer, and so, this approach significantly\nignores the role of the answer. In this paper, we devise a novel\nimage-question-answer synergistic network to value the role of the answer for\nprecise visual dialog. We extend the traditional one-stage solution to a\ntwo-stage solution. In the first stage, candidate answers are coarsely scored\naccording to their relevance to the image and question pair. Afterward, in the\nsecond stage, answers with high probability of being correct are re-ranked by\nsynergizing with image and question. On the Visual Dialog v1.0 dataset, the\nproposed synergistic network boosts the discriminative visual dialog model to\nachieve a new state-of-the-art of 57.88\\% normalized discounted cumulative\ngain. A generative visual dialog model equipped with the proposed technique\nalso shows promising improvements.","url_abs":"http://arxiv.org/abs/1902.09774v1","url_pdf":"http://arxiv.org/pdf/1902.09774v1.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":[],"tasks":[{"task_slug":"visual-dialogue","task_name":"Visual Dialog"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-dialog-on-visual-dialog-v1-0-test-std","task":"Visual Dialog","dataset":"Visual Dialog v1.0 test-std","model":"Synergistic","rank_in_archive_order":57,"of":80,"metrics":{"MRR (x 100)":"62.20","Mean":"4.17","NDCG (x 100)":"57.32","R@1":"47.90","R@5":"80.43"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.09774","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}