{"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/an-empirical-study-and-analysis-of","title":"An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild","arxiv_id":"1605.04253","date":"2016-05-13","proceeding":null,"authors":["Wei-Lun Chao","Soravit Changpinyo","Boqing Gong","Fei Sha"],"abstract":"Zero-shot learning (ZSL) methods have been studied in the unrealistic setting\nwhere test data are assumed to come from unseen classes only. In this paper, we\nadvocate studying the problem of generalized zero-shot learning (GZSL) where\nthe test data's class memberships are unconstrained. We show empirically that\nnaively using the classifiers constructed by ZSL approaches does not perform\nwell in the generalized setting. Motivated by this, we propose a simple but\neffective calibration method that can be used to balance two conflicting\nforces: recognizing data from seen classes versus those from unseen ones. We\ndevelop a performance metric to characterize such a trade-off and examine the\nutility of this metric in evaluating various ZSL approaches. Our analysis\nfurther shows that there is a large gap between the performance of existing\napproaches and an upper bound established via idealized semantic embeddings,\nsuggesting that improving class semantic embeddings is vital to GZSL.","url_abs":"http://arxiv.org/abs/1605.04253v2","url_pdf":"http://arxiv.org/pdf/1605.04253v2.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":"an-empirical-study-and-analysis-of","repo_url":"https://github.com/pujols/Zero-shot-learning-journal","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"generalized-zero-shot-learning","task_name":"Generalized Zero-Shot Learning"},{"task_slug":"object-recognition","task_name":"Object Recognition"},{"task_slug":"zero-shot-learning","task_name":"Zero-Shot Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1605.04253","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}