{"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/unsupervised-visual-sense-disambiguation-for","title":"Unsupervised Visual Sense Disambiguation for Verbs using Multimodal Embeddings","arxiv_id":"1603.09188","date":"2016-03-30","proceeding":"NAACL 2016 6","authors":["Spandana Gella","Mirella Lapata","Frank Keller"],"abstract":"We introduce a new task, visual sense disambiguation for verbs: given an\nimage and a verb, assign the correct sense of the verb, i.e., the one that\ndescribes the action depicted in the image. Just as textual word sense\ndisambiguation is useful for a wide range of NLP tasks, visual sense\ndisambiguation can be useful for multimodal tasks such as image retrieval,\nimage description, and text illustration. We introduce VerSe, a new dataset\nthat augments existing multimodal datasets (COCO and TUHOI) with sense labels.\nWe propose an unsupervised algorithm based on Lesk which performs visual sense\ndisambiguation using textual, visual, or multimodal embeddings. We find that\ntextual embeddings perform well when gold-standard textual annotations (object\nlabels and image descriptions) are available, while multimodal embeddings\nperform well on unannotated images. We also verify our findings by using the\ntextual and multimodal embeddings as features in a supervised setting and\nanalyse the performance of visual sense disambiguation task. VerSe is made\npublicly available and can be downloaded at:\nhttps://github.com/spandanagella/verse.","url_abs":"http://arxiv.org/abs/1603.09188v1","url_pdf":"http://arxiv.org/pdf/1603.09188v1.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":"unsupervised-visual-sense-disambiguation-for","repo_url":"https://github.com/spandanagella/verse","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":null,"task_name":"Image Description"},{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"word-sense-disambiguation","task_name":"Word Sense Disambiguation"}],"methods":[],"datasets_introduced":[{"slug":"verse","name":"Verse","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1603.09188","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}