{"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/learning-like-a-child-fast-novel-visual","title":"Learning like a Child: Fast Novel Visual Concept Learning from Sentence Descriptions of Images","arxiv_id":"1504.06692","date":"2015-04-25","proceeding":"ICCV 2015 12","authors":["Junhua Mao","Wei Xu","Yi Yang","Jiang Wang","Zhiheng Huang","Alan Yuille"],"abstract":"In this paper, we address the task of learning novel visual concepts, and\ntheir interactions with other concepts, from a few images with sentence\ndescriptions. Using linguistic context and visual features, our method is able\nto efficiently hypothesize the semantic meaning of new words and add them to\nits word dictionary so that they can be used to describe images which contain\nthese novel concepts. Our method has an image captioning module based on m-RNN\nwith several improvements. In particular, we propose a transposed weight\nsharing scheme, which not only improves performance on image captioning, but\nalso makes the model more suitable for the novel concept learning task. We\npropose methods to prevent overfitting the new concepts. In addition, three\nnovel concept datasets are constructed for this new task. In the experiments,\nwe show that our method effectively learns novel visual concepts from a few\nexamples without disturbing the previously learned concepts. The project page\nis http://www.stat.ucla.edu/~junhua.mao/projects/child_learning.html","url_abs":"http://arxiv.org/abs/1504.06692v2","url_pdf":"http://arxiv.org/pdf/1504.06692v2.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":"learning-like-a-child-fast-novel-visual","repo_url":"https://github.com/mjhucla/TF-mRNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"image-captioning","task_name":"Image Captioning"},{"task_slug":"novel-concepts","task_name":"Novel Concepts"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}