Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images

No-reference image quality assessment (NR-IQA), which devotes to predicting image quality without relying on the corresponding pristine counterpart, develops rapidly in recent years. However, little investigation has been dedicated to quality assessment of realistic night-time images. Existing NR-IQ...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Bowen Li, Xianpei Wang, Weixia Zhang, Meng Tian, Hongtai Yao
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2020
Materias:	No-reference image quality assessment convolutional neural network night-time images multi-stream CNN dual head network Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/e7531f9ed5ec4f0d81f9c0c614e0e566
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:e7531f9ed5ec4f0d81f9c0c614e0e566
record_format	dspace
spelling	oai:doaj.org-article:e7531f9ed5ec4f0d81f9c0c614e0e5662021-11-19T00:05:22ZDual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images2169-353610.1109/ACCESS.2020.3020750https://doaj.org/article/e7531f9ed5ec4f0d81f9c0c614e0e5662020-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9181551/https://doaj.org/toc/2169-3536No-reference image quality assessment (NR-IQA), which devotes to predicting image quality without relying on the corresponding pristine counterpart, develops rapidly in recent years. However, little investigation has been dedicated to quality assessment of realistic night-time images. Existing NR-IQA algorithms laboriously cope with this night-time scenario since complicated authentic distortions such as low contrast, blurred details, and reduced visibility usually appear on it. In this paper, we propose an end-to-end NR-IQA model to meet this challenge based on a multi-stream deep convolutional neural network (DCNN). Two streams, brightness-aware CNN and naturalness-aware CNN are constructed respectively by a brightness-altered image identification task with a self-established dataset and a quality-prediction regression task with an existing authentically-distorted IQA dataset to improve quality-aware initializations. In this case, given the quick convergence and little transformation in the lower layers, a shallow-layer-shared architecture is explored to reduce computational cost. Finally, the features of these two pipelines are collected by an effective pooling method and then concatenated as the image representation for fine-tuning. The effectiveness and efficiency of the proposed method are verified by several different experiments on the NNID, CCRIQ and LIVE Challenge databases. Furthermore, the superiority of wide applications such as for contrast-distorted and driving scenarios is demonstrated on the CID2013, CCID2014 and BBD-100k databases.Bowen LiXianpei WangWeixia ZhangMeng TianHongtai YaoIEEEarticleNo-reference image quality assessmentconvolutional neural networknight-time imagesmulti-stream CNNdual head networkElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 8, Pp 158585-158599 (2020)
institution	DOAJ
collection	DOAJ
language	EN
topic	No-reference image quality assessment convolutional neural network night-time images multi-stream CNN dual head network Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	No-reference image quality assessment convolutional neural network night-time images multi-stream CNN dual head network Electrical engineering. Electronics. Nuclear engineering TK1-9971 Bowen Li Xianpei Wang Weixia Zhang Meng Tian Hongtai Yao Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
description	No-reference image quality assessment (NR-IQA), which devotes to predicting image quality without relying on the corresponding pristine counterpart, develops rapidly in recent years. However, little investigation has been dedicated to quality assessment of realistic night-time images. Existing NR-IQA algorithms laboriously cope with this night-time scenario since complicated authentic distortions such as low contrast, blurred details, and reduced visibility usually appear on it. In this paper, we propose an end-to-end NR-IQA model to meet this challenge based on a multi-stream deep convolutional neural network (DCNN). Two streams, brightness-aware CNN and naturalness-aware CNN are constructed respectively by a brightness-altered image identification task with a self-established dataset and a quality-prediction regression task with an existing authentically-distorted IQA dataset to improve quality-aware initializations. In this case, given the quick convergence and little transformation in the lower layers, a shallow-layer-shared architecture is explored to reduce computational cost. Finally, the features of these two pipelines are collected by an effective pooling method and then concatenated as the image representation for fine-tuning. The effectiveness and efficiency of the proposed method are verified by several different experiments on the NNID, CCRIQ and LIVE Challenge databases. Furthermore, the superiority of wide applications such as for contrast-distorted and driving scenarios is demonstrated on the CID2013, CCID2014 and BBD-100k databases.
format	article
author	Bowen Li Xianpei Wang Weixia Zhang Meng Tian Hongtai Yao
author_facet	Bowen Li Xianpei Wang Weixia Zhang Meng Tian Hongtai Yao
author_sort	Bowen Li
title	Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
title_short	Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
title_full	Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
title_fullStr	Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
title_full_unstemmed	Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images
title_sort	dual head network for no-reference quality assessment towards realistic night-time images
publisher	IEEE
publishDate	2020
url	https://doaj.org/article/e7531f9ed5ec4f0d81f9c0c614e0e566
work_keys_str_mv	AT bowenli dualheadnetworkfornoreferencequalityassessmenttowardsrealisticnighttimeimages AT xianpeiwang dualheadnetworkfornoreferencequalityassessmenttowardsrealisticnighttimeimages AT weixiazhang dualheadnetworkfornoreferencequalityassessmenttowardsrealisticnighttimeimages AT mengtian dualheadnetworkfornoreferencequalityassessmenttowardsrealisticnighttimeimages AT hongtaiyao dualheadnetworkfornoreferencequalityassessmenttowardsrealisticnighttimeimages
_version_	1718420702177001472

Dual Head Network for No-Reference Quality Assessment Towards Realistic Night-Time Images

Ejemplares similares