{"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/deep-cross-residual-learning-for-multitask","title":"Deep Cross Residual Learning for Multitask Visual Recognition","arxiv_id":"1604.01335","date":"2016-04-05","proceeding":null,"authors":["Brendan Jou","Shih-Fu Chang"],"abstract":"Residual learning has recently surfaced as an effective means of constructing\nvery deep neural networks for object recognition. However, current incarnations\nof residual networks do not allow for the modeling and integration of complex\nrelations between closely coupled recognition tasks or across domains. Such\nproblems are often encountered in multimedia applications involving large-scale\ncontent recognition. We propose a novel extension of residual learning for deep\nnetworks that enables intuitive learning across multiple related tasks using\ncross-connections called cross-residuals. These cross-residuals connections can\nbe viewed as a form of in-network regularization and enables greater network\ngeneralization. We show how cross-residual learning (CRL) can be integrated in\nmultitask networks to jointly train and detect visual concepts across several\ntasks. We present a single multitask cross-residual network with >40% less\nparameters that is able to achieve competitive, or even better, detection\nperformance on a visual sentiment concept detection problem normally requiring\nmultiple specialized single-task networks. The resulting multitask\ncross-residual network also achieves better detection performance by about\n10.4% over a standard multitask residual network without cross-residuals with\neven a small amount of cross-task weighting.","url_abs":"http://arxiv.org/abs/1604.01335v2","url_pdf":"http://arxiv.org/pdf/1604.01335v2.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":"deep-cross-residual-learning-for-multitask","repo_url":"https://github.com/imatge-upc/affective-2017-musa2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"object-recognition","task_name":"Object Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1604.01335","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}