Discriminative convolutional neural network for image quality assessment with fixed convolution filters

Motohiro Takagi, Akito Sakurai, Masafumi Hagiwara

Research output: Contribution to journalArticle

Abstract

Current image quality assessment (IQA) methods require the original images for evaluation. However, recently, IQA methods that use machine learning have been proposed. These methods learn the relationship between the distorted image and the image quality automatically. In this paper, we propose an IQA method based on deep learning that does not require a reference image. We show that a convolutional neural network with distortion prediction and fixed filters improves the IQA accuracy.

Original languageEnglish
Pages (from-to)2265-2266
Number of pages2
JournalIEICE Transactions on Information and Systems
VolumeE102D
Issue number11
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Convolution
Image quality
Neural networks
Learning systems

Keywords

  • Convolutional neural network
  • Image quality assessment
  • Subjective image quality

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Cite this

Discriminative convolutional neural network for image quality assessment with fixed convolution filters. / Takagi, Motohiro; Sakurai, Akito; Hagiwara, Masafumi.

In: IEICE Transactions on Information and Systems, Vol. E102D, No. 11, 01.01.2019, p. 2265-2266.

Research output: Contribution to journalArticle

@article{a1f02cf3412d449982f4601b61f28e9e,
title = "Discriminative convolutional neural network for image quality assessment with fixed convolution filters",
abstract = "Current image quality assessment (IQA) methods require the original images for evaluation. However, recently, IQA methods that use machine learning have been proposed. These methods learn the relationship between the distorted image and the image quality automatically. In this paper, we propose an IQA method based on deep learning that does not require a reference image. We show that a convolutional neural network with distortion prediction and fixed filters improves the IQA accuracy.",
keywords = "Convolutional neural network, Image quality assessment, Subjective image quality",
author = "Motohiro Takagi and Akito Sakurai and Masafumi Hagiwara",
year = "2019",
month = "1",
day = "1",
doi = "10.1587/transinf.2018EDL8272",
language = "English",
volume = "E102D",
pages = "2265--2266",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "11",

}

TY - JOUR

T1 - Discriminative convolutional neural network for image quality assessment with fixed convolution filters

AU - Takagi, Motohiro

AU - Sakurai, Akito

AU - Hagiwara, Masafumi

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Current image quality assessment (IQA) methods require the original images for evaluation. However, recently, IQA methods that use machine learning have been proposed. These methods learn the relationship between the distorted image and the image quality automatically. In this paper, we propose an IQA method based on deep learning that does not require a reference image. We show that a convolutional neural network with distortion prediction and fixed filters improves the IQA accuracy.

AB - Current image quality assessment (IQA) methods require the original images for evaluation. However, recently, IQA methods that use machine learning have been proposed. These methods learn the relationship between the distorted image and the image quality automatically. In this paper, we propose an IQA method based on deep learning that does not require a reference image. We show that a convolutional neural network with distortion prediction and fixed filters improves the IQA accuracy.

KW - Convolutional neural network

KW - Image quality assessment

KW - Subjective image quality

UR - http://www.scopus.com/inward/record.url?scp=85075425689&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075425689&partnerID=8YFLogxK

U2 - 10.1587/transinf.2018EDL8272

DO - 10.1587/transinf.2018EDL8272

M3 - Article

AN - SCOPUS:85075425689

VL - E102D

SP - 2265

EP - 2266

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 11

ER -