Blind Image Quality Assessment via Deep Learning

CORC > 西安光学精密机械研究所 > 中国科学院西安光学精密机械研究所 > 光学影像学习与分析中心

	Blind Image Quality Assessment via Deep Learning
	Hou, Weilong 1; Gao, Xinbo 1; Tao, Dacheng2,3 ; Li, Xuelong4
刊名	ieee transactions on neural networks and learning systems
	2015-06-01
卷号	26 期号:6 页码:1275-1286
关键词	Deep learning image quality assessment (IQA) natural scene statistics (NSS) no reference
英文摘要	this paper investigates how to blindly evaluate the visual quality of an image by learning rules from linguistic descriptions. extensive psychological evidence shows that humans prefer to conduct evaluations qualitatively rather than numerically. the qualitative evaluations are then converted into the numerical scores to fairly benchmark objective image quality assessment (iqa) metrics. recently, lots of learning-based iqa models are proposed by analyzing the mapping from the images to numerical ratings. however, the learnt mapping can hardly be accurate enough because some information has been lost in such an irreversible conversion from the linguistic descriptions to numerical scores. in this paper, we propose a blind iqa model, which learns qualitative evaluations directly and outputs numerical scores for general utilization and fair comparison. images are represented by natural scene statistics features. a discriminative deep model is trained to classify the features into five grades, corresponding to five explicit mental concepts, i.e., excellent, good, fair, poor, and bad. a newly designed quality pooling is then applied to convert the qualitative labels into scores. the classification framework is not only much more natural than the regression-based models, but also robust to the small sample size problem. thorough experiments are conducted on popular databases to verify the model's effectiveness, efficiency, and robustness.
WOS标题词	science & technology ; technology
类目[WOS]	computer science, artificial intelligence ; computer science, hardware & architecture ; computer science, theory & methods ; engineering, electrical & electronic
研究领域[WOS]	computer science ; engineering
关键词[WOS]	natural scene statistics ; neural-network ; domain ; representation ; information ; framework
收录类别	SCI ; EI
语种	英语
WOS记录号	WOS:000354957000013
公开日期	2015-07-28
内容类型	期刊论文
源URL	[http://ir.opt.ac.cn/handle/181661/25058]
专题	西安光学精密机械研究所_光学影像学习与分析中心
作者单位	1.Xidian Univ, Sch Elect Engn, Xian 710071, Shaanxi, Peoples R China 2.Univ Technol Sydney, Ctr Quantum Computat & Intelligent Syst, Ultimo, NSW 2007, Australia 3.Univ Technol Sydney, Fac Engn & Informat Technol, Ultimo, NSW 2007, Australia 4.Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Ctr Opt IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China
推荐引用方式 GB/T 7714	Hou, Weilong,Gao, Xinbo,Tao, Dacheng,et al. Blind Image Quality Assessment via Deep Learning[J]. ieee transactions on neural networks and learning systems,2015,26(6):1275-1286.
APA	Hou, Weilong,Gao, Xinbo,Tao, Dacheng,&Li, Xuelong.(2015).Blind Image Quality Assessment via Deep Learning.ieee transactions on neural networks and learning systems,26(6),1275-1286.
MLA	Hou, Weilong,et al."Blind Image Quality Assessment via Deep Learning".ieee transactions on neural networks and learning systems 26.6(2015):1275-1286.