{"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/topic-guided-attention-for-image-captioning","title":"Topic-Guided Attention for Image Captioning","arxiv_id":"1807.03514","date":"2018-07-10","proceeding":null,"authors":["Zhihao Zhu","Zhan Xue","Zejian yuan"],"abstract":"Attention mechanisms have attracted considerable interest in image captioning\nbecause of its powerful performance. Existing attention-based models use\nfeedback information from the caption generator as guidance to determine which\nof the image features should be attended to. A common defect of these attention\ngeneration methods is that they lack a higher-level guiding information from\nthe image itself, which sets a limit on selecting the most informative image\nfeatures. Therefore, in this paper, we propose a novel attention mechanism,\ncalled topic-guided attention, which integrates image topics in the attention\nmodel as a guiding information to help select the most important image\nfeatures. Moreover, we extract image features and image topics with separate\nnetworks, which can be fine-tuned jointly in an end-to-end manner during\ntraining. The experimental results on the benchmark Microsoft COCO dataset show\nthat our method yields state-of-art performance on various quantitative\nmetrics.","url_abs":"http://arxiv.org/abs/1807.03514v1","url_pdf":"http://arxiv.org/pdf/1807.03514v1.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":"topic-guided-attention-for-image-captioning","repo_url":"https://github.com/jsaikmr/Building-a-Topic-Modeling-for-Images-using-LDA-and-Transfer-Learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"image-captioning","task_name":"Image Captioning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1807.03514","atlas_url":"https://app.syntology.ai/?focus=1807.03514","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}