{"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-neural-compositional-paradigm-for-image","title":"A Neural Compositional Paradigm for Image Captioning","arxiv_id":"1810.09630","date":"2018-10-23","proceeding":"NeurIPS 2018 12","authors":["Bo Dai","Sanja Fidler","Dahua Lin"],"abstract":"Mainstream captioning models often follow a sequential structure to generate\ncaptions, leading to issues such as introduction of irrelevant semantics, lack\nof diversity in the generated captions, and inadequate generalization\nperformance. In this paper, we present an alternative paradigm for image\ncaptioning, which factorizes the captioning procedure into two stages: (1)\nextracting an explicit semantic representation from the given image; and (2)\nconstructing the caption based on a recursive compositional procedure in a\nbottom-up manner. Compared to conventional ones, our paradigm better preserves\nthe semantic content through an explicit factorization of semantics and syntax.\nBy using the compositional generation procedure, caption construction follows a\nrecursive structure, which naturally fits the properties of human language.\nMoreover, the proposed compositional procedure requires less data to train,\ngeneralizes better, and yields more diverse captions.","url_abs":"http://arxiv.org/abs/1810.09630v1","url_pdf":"http://arxiv.org/pdf/1810.09630v1.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-neural-compositional-paradigm-for-image","repo_url":"https://github.com/ajaysub110/A-Neural-Compositional-Paradigm-for-Image-Captioning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"diversity","task_name":"Diversity"},{"task_slug":"image-captioning","task_name":"Image Captioning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1810.09630","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}