Not All Samples are Trustworthy: Towards Deep Robust SVP Prediction

doi:10.1109/TPAMI.2020.3047817

	Not All Samples are Trustworthy: Towards Deep Robust SVP Prediction
	Xu, Qianqian 1; Yang, Zhiyong 2,3; Jiang, Yangbangyan 2,3; Cao, Xiaochun 2,3,4; Yao, Yuan 5,6; Huang, Qingming 1,7,8,9
	2022-06-01
发表期刊	IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
ISSN	0162-8828
卷号	44 期号:6 页码:3154-3169
摘要	In this paper, we study the problem of estimating subjective visual properties (SVP) for images, which is an emerging task in Computer Vision. Generally speaking, collecting SVP datasets involves a crowdsourcing process where annotations are obtained from a wide range of online users. Since the process is done without quality control, SVP datasets are known to suffer from noise. This leads to the issue that not all samples are trustworthy. Facing this problem, we need to develop robust models for learning SVP from noisy crowdsourced annotations. In this paper, we construct two general robust learning frameworks for this application. Specifically, in the first framework, we propose a probabilistic framework to explicitly model the sparse unreliable patterns that exist in the dataset. It is noteworthy that we then provide an alternative framework that could reformulate the sparse unreliable patterns as a "contraction" operation over the original loss function. The latter framework leverages not only efficient end-to-end training but also rigorous theoretical analyses. To apply these frameworks, we further provide two models as implementations of the frameworks, where the sparse noise parameters could be interpreted with the HodgeRank theory. Finally, extensive theoretical and empirical studies show the effectiveness of our proposed framework.
关键词	Noise measurement Annotations Task analysis Predictive models Robustness Visualization Training Subjective visual property (SVP) robustness outlier detection probabilistic model
DOI	10.1109/TPAMI.2020.3047817
收录类别	SCI
语种	英语
资助项目	National Key R&D Program of China[2018AAA0102003] ; Key Research Program of Frontier Sciences, CAS[QYZDJ-SSW-SYS013] ; Beijing Education Committee Cooperation Beijing Natural Science Foundation ; Youth Innovation Promotion Association CAS ; Strategic Priority Research Program of Chinese Academy of Sciences[XDB28000000] ; Hong Kong Research Grant Council (HKRGC)[16303817] ; Hong Kong Research Grant Council (HKRGC)[ITF UIM/390] ; Tencent AI Lab, Si Family Foundation ; National Natural Science Foundation of China[61620106009] ; National Natural Science Foundation of China[U1936208] ; National Natural Science Foundation of China[61733007] ; National Natural Science Foundation of China[U1736219] ; National Natural Science Foundation of China[61931008] ; National Natural Science Foundation of China[61976202] ; National Natural Science Foundation of China[61836002] ; Tencent AI Lab ; Microsoft Research-Asia
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000803117500027
出版者	IEEE COMPUTER SOC
引用统计	被引频次：2[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/19575
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Huang, Qingming
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 2.Chinese Acad Sci, Inst Informat Engn, State Key Lab Informat Secur, Beijing 100093, Peoples R China 3.Univ Chinese Acad Sci, Sch Cyber Secur, Beijing 100049, Peoples R China 4.Peng Cheng Lab, Cyberspace Secur Res Ctr, Shenzhen 518055, Peoples R China 5.Hong Kong Univ Sci & Technol, Dept Math, Hong Kong, Peoples R China 6.Hong Kong Univ Sci & Technol, Courtesy Comp Sci & Engn, Hong Kong, Peoples R China 7.Univ Chinese Acad Sci, Sch Comp Sci & Technol, Beijing 101408, Peoples R China 8.Univ Chinese Acad Sci, Key Lab Big Data Min & Knowledge Management BDKM, Beijing 101408, Peoples R China 9.Peng Cheng Lab, Shenzhen 518055, Peoples R China
推荐引用方式 GB/T 7714	Xu, Qianqian,Yang, Zhiyong,Jiang, Yangbangyan,et al. Not All Samples are Trustworthy: Towards Deep Robust SVP Prediction[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,2022,44(6):3154-3169.
APA	Xu, Qianqian,Yang, Zhiyong,Jiang, Yangbangyan,Cao, Xiaochun,Yao, Yuan,&Huang, Qingming.(2022).Not All Samples are Trustworthy: Towards Deep Robust SVP Prediction.IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,44(6),3154-3169.
MLA	Xu, Qianqian,et al."Not All Samples are Trustworthy: Towards Deep Robust SVP Prediction".IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 44.6(2022):3154-3169.