DeepSim: Deep similarity for image quality assessment

被引：131

作者：

Gao, Fei ^{[1
]}

Wang, Yi ^{[2
]}

Li, Panpeng ^{[1
]}

Tan, Min ^{[1
]}

Yu, Jun ^{[1
]}

Zhu, Yani ^{[3
,4
]}

机构：

[1] Hangzhou Dianzi Univ, Key Lab Complex Syst Modeling & Simulat, Sch Comp Sci & Technol, Hangzhou, Zhejiang, Peoples R China

[2] Zhejiang Univ City Coll, Dept Visual Commun Design, Artist Design & Creat Sch, Hangzhou, Zhejiang, Peoples R China

[3] Zhejiang Univ Technol, Coll Informat Engn, Hangzhou, Zhejiang, Peoples R China

[4] Hangzhou Dianzi Univ, Inst Sci & Technol, Hangzhou, Zhejiang, Peoples R China

来源：

NEUROCOMPUTING | 2017年 / 257卷

基金：

中国国家自然科学基金;

关键词：

Image quality assessment; Deep learning; Pooling; Structural similarity; Convolutional neural networks (CNN); NATURAL SCENE STATISTICS;

D O I：

10.1016/j.neucom.2017.01.054

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper studies one interesting problem: how does the deep neural network (DNN) architecture affect the image quality assessment (IQA) performance? In order to find the answer, we propose a novel full reference IQA framework, codenamed deep similarity (DeepSim). In DeepSim, we first measure the local similarities between the features (produced by a DNN model) of the test image and those of the reference image; Afterwards, the local quality indices are gradually pooled together to estimate the overall quality score. In addition; various factors that may affect the IQA performance are investigated. Thorough experiments conducted on standard databases show that: (1) DeepSim can accurately predict human perceived image quality and outperforms previous state-of-the-art; (2) mid-level representations are most effective for quality prediction; and (3) preprocessing, the restricted linear units and max-pooling operations are beneficial for the IQA performance. (C) 2017 Elsevier B.V. All rights reserved.

引用

页码：104 / 114

页数：11

共 44 条

[1] No-Reference Blur Assessment of Digital Pictures Based on Multifeature Classifiers [J].