Adversarial Learning Guided Task Relatedness Refinement for Multi-Task Deep Learning

doi:10.1109/TMM.2022.3216460

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	Adversarial Learning Guided Task Relatedness Refinement for Multi-Task Deep Learning
	Fang, Yuchun 1; Cai, Sirui 1; Cao, Yiting 1; Li, Zhengchen 1; Zhang, Zhaoxiang 2,3
刊名	IEEE TRANSACTIONS ON MULTIMEDIA
	2023
卷号	25 页码:6946-6957
关键词	Index Terms-Multi-task learning deep learning task relatedness
ISSN号	1520-9210
DOI	10.1109/TMM.2022.3216460
通讯作者	Fang, Yuchun(ycfang@shu.edu.cn)
英文摘要	In machine learning, the relatedness across multiple tasks is usually complex and entangled. Due to dataset bias, the relatedness among tasks might be distorted and mislead the training of the models with solid learning ability, such as the multi-task neural networks. In this paper, we propose the idea of Relatedness Refinement Multi-Task Learning (RRMTDL) by introducing adversarial learning in the multi-task deep neural network to tackle the problem. The RRMTDL deep learning model restrains the misleading relatedness task by adversarial training and extracts information sharing across tasks with valuable relatedness. With RRMTDL, multi-task deep learning can enhance the task-specific representation for the major tasks by excluding the misleading relatedness. We design tests with various combinations of task-relatedness to validate the proposed model. Experimental results show that the RRMTDL model can effectively refine the task relatedness and prominently outperform other multi-task deep learning models in datasets with entangled task labels.
资助项目	National Natural Science Foundation of China[61976132] ; National Natural Science Foundation of China[61991411] ; National Natural Science Foundation of China[U1811461] ; Natural Science Foundation of Shanghai[19ZR1419200]
WOS研究方向	Computer Science ; Telecommunications
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001102654000017
资助机构	National Natural Science Foundation of China ; Natural Science Foundation of Shanghai
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/55155]
专题	多模态人工智能系统全国重点实验室
通讯作者	Fang, Yuchun
作者单位	1.Shanghai Univ, Sch Comp Engn & Sci, Shanghai 200444, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 3.HKISI CAS, Ctr Artificial Intelligence & Robot, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Fang, Yuchun,Cai, Sirui,Cao, Yiting,et al. Adversarial Learning Guided Task Relatedness Refinement for Multi-Task Deep Learning[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2023,25:6946-6957.
APA	Fang, Yuchun,Cai, Sirui,Cao, Yiting,Li, Zhengchen,&Zhang, Zhaoxiang.(2023).Adversarial Learning Guided Task Relatedness Refinement for Multi-Task Deep Learning.IEEE TRANSACTIONS ON MULTIMEDIA,25,6946-6957.
MLA	Fang, Yuchun,et al."Adversarial Learning Guided Task Relatedness Refinement for Multi-Task Deep Learning".IEEE TRANSACTIONS ON MULTIMEDIA 25(2023):6946-6957.