Semantic-aware multi-task learning for image aesthetic quality assessment

In recent years, image aesthetic quality assessment has attracted considerable attention due to the massive growth of digital images in social platforms and the Internet. However, automatically assessing aesthetic quality of an image is a challenging task, because image aesthetic is affected by vari...

Full description

Saved in:
Bibliographic Details
Main Authors: Weiliang Yan, Yuqing Li, Huan Yang, Baoxiang Huang, Zhenkuan Pan
Format: Article
Language:English
Published: Taylor & Francis Group 2022-12-01
Series:Connection Science
Subjects:
Online Access:http://dx.doi.org/10.1080/09540091.2022.2147902
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850168165662720000
author Weiliang Yan
Yuqing Li
Huan Yang
Baoxiang Huang
Zhenkuan Pan
author_facet Weiliang Yan
Yuqing Li
Huan Yang
Baoxiang Huang
Zhenkuan Pan
author_sort Weiliang Yan
collection DOAJ
description In recent years, image aesthetic quality assessment has attracted considerable attention due to the massive growth of digital images in social platforms and the Internet. However, automatically assessing aesthetic quality of an image is a challenging task, because image aesthetic is affected by various factors, and the criteria for judging the aesthetic of images with diverse semantic information are different. To this end, a Semantic-Aware Multi-task convolution neural network (SAM-CNN) for evaluating image aesthetic quality is proposed in this paper. The network can fuse intermediate features of different layers at different scales in CNN to obtain a more comprehensive and accurate aesthetic expression, under the joint supervision of image aesthetic quality assessment task and semantic classification task in a multi-task learning manner. Besides, by applying the attention mechanism, semantic information with a large receptive field extracted from deep layers is utilised to guide the network to focus on the key parts of features to be fused, to improve the effectiveness of feature fusion. Experimental results on the AVA dataset and Photo.net dataset demonstrate the effectiveness and superiority of the proposed SAM-CNN.
format Article
id doaj-art-e61998bd3ba24d84a01f5b6263f7ed13
institution OA Journals
issn 0954-0091
1360-0494
language English
publishDate 2022-12-01
publisher Taylor & Francis Group
record_format Article
series Connection Science
spelling doaj-art-e61998bd3ba24d84a01f5b6263f7ed132025-08-20T02:21:02ZengTaylor & Francis GroupConnection Science0954-00911360-04942022-12-013412689271310.1080/09540091.2022.21479022147902Semantic-aware multi-task learning for image aesthetic quality assessmentWeiliang Yan0Yuqing Li1Huan Yang2Baoxiang Huang3Zhenkuan Pan4College of Computer Science and Technology, Qingdao UniversityCollege of Computer Science and Technology, Qingdao UniversityCollege of Computer Science and Technology, Qingdao UniversityCollege of Computer Science and Technology, Qingdao UniversityCollege of Computer Science and Technology, Qingdao UniversityIn recent years, image aesthetic quality assessment has attracted considerable attention due to the massive growth of digital images in social platforms and the Internet. However, automatically assessing aesthetic quality of an image is a challenging task, because image aesthetic is affected by various factors, and the criteria for judging the aesthetic of images with diverse semantic information are different. To this end, a Semantic-Aware Multi-task convolution neural network (SAM-CNN) for evaluating image aesthetic quality is proposed in this paper. The network can fuse intermediate features of different layers at different scales in CNN to obtain a more comprehensive and accurate aesthetic expression, under the joint supervision of image aesthetic quality assessment task and semantic classification task in a multi-task learning manner. Besides, by applying the attention mechanism, semantic information with a large receptive field extracted from deep layers is utilised to guide the network to focus on the key parts of features to be fused, to improve the effectiveness of feature fusion. Experimental results on the AVA dataset and Photo.net dataset demonstrate the effectiveness and superiority of the proposed SAM-CNN.http://dx.doi.org/10.1080/09540091.2022.2147902image aesthetic quality assessmentdeep learningfeature fusionattention mechanismsemantic aware
spellingShingle Weiliang Yan
Yuqing Li
Huan Yang
Baoxiang Huang
Zhenkuan Pan
Semantic-aware multi-task learning for image aesthetic quality assessment
Connection Science
image aesthetic quality assessment
deep learning
feature fusion
attention mechanism
semantic aware
title Semantic-aware multi-task learning for image aesthetic quality assessment
title_full Semantic-aware multi-task learning for image aesthetic quality assessment
title_fullStr Semantic-aware multi-task learning for image aesthetic quality assessment
title_full_unstemmed Semantic-aware multi-task learning for image aesthetic quality assessment
title_short Semantic-aware multi-task learning for image aesthetic quality assessment
title_sort semantic aware multi task learning for image aesthetic quality assessment
topic image aesthetic quality assessment
deep learning
feature fusion
attention mechanism
semantic aware
url http://dx.doi.org/10.1080/09540091.2022.2147902
work_keys_str_mv AT weiliangyan semanticawaremultitasklearningforimageaestheticqualityassessment
AT yuqingli semanticawaremultitasklearningforimageaestheticqualityassessment
AT huanyang semanticawaremultitasklearningforimageaestheticqualityassessment
AT baoxianghuang semanticawaremultitasklearningforimageaestheticqualityassessment
AT zhenkuanpan semanticawaremultitasklearningforimageaestheticqualityassessment