{"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/fully-adaptive-feature-sharing-in-multi-task","title":"Fully-adaptive Feature Sharing in Multi-Task Networks with Applications in Person Attribute Classification","arxiv_id":"1611.05377","date":"2016-11-16","proceeding":"CVPR 2017 7","authors":["Yongxi Lu","Abhishek Kumar","Shuangfei Zhai","Yu Cheng","Tara Javidi","Rogerio Feris"],"abstract":"Multi-task learning aims to improve generalization performance of multiple\nprediction tasks by appropriately sharing relevant information across them. In\nthe context of deep neural networks, this idea is often realized by\nhand-designed network architectures with layers that are shared across tasks\nand branches that encode task-specific features. However, the space of possible\nmulti-task deep architectures is combinatorially large and often the final\narchitecture is arrived at by manual exploration of this space subject to\ndesigner's bias, which can be both error-prone and tedious. In this work, we\npropose a principled approach for designing compact multi-task deep learning\narchitectures. Our approach starts with a thin network and dynamically widens\nit in a greedy manner during training using a novel criterion that promotes\ngrouping of similar tasks together. Our Extensive evaluation on person\nattributes classification tasks involving facial and clothing attributes\nsuggests that the models produced by the proposed method are fast, compact and\ncan closely match or exceed the state-of-the-art accuracy from strong baselines\nby much more expensive models.","url_abs":"http://arxiv.org/abs/1611.05377v1","url_pdf":"http://arxiv.org/pdf/1611.05377v1.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":"fully-adaptive-feature-sharing-in-multi-task","repo_url":"https://github.com/luyongxi/deep_share","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"caffe2","reach":{"status":"ok"}}],"tasks":[{"task_slug":"attribute","task_name":"Attribute"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.05377","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}