Property-specific aesthetic assessment with unsupervised aesthetic property discovery

Jun Tae Lee; Chul Lee; Chang Su Kim

doi:10.1109/ACCESS.2019.2936289

Property-specific aesthetic assessment with unsupervised aesthetic property discovery

Jun Tae Lee, Chul Lee, Chang Su Kim^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

14 Citations (Scopus)

Abstract

We propose the property-specific aesthetic assessment (PSAA) algorithm with unsupervised aesthetic property discovery. The proposed PSAA algorithm uses an aesthetic feature extractor, an aesthetic property classifier, and multiple property-specific assessment networks. The aesthetic feature extractor analyzes aesthetics of images to generate features. Using such aesthetic features, we discover diverse aesthetic properties in an unsupervised manner and develop the aesthetic property classifier to predict the aesthetic property of each image. For each discovered aesthetic property, we train a property-specific assessment network. Thus, we can assess the aesthetic quality of an image using the property-specific network that corresponds to its property. Experimental results on a large dataset show that the proposed PSAA algorithm achieves state-of-the-art aesthetic assessment performance. Furthermore, we demonstrate that PSAA is useful for improving aesthetic qualities of images in two applications: contrast enhancement and image cropping.

Original language	English
Article number	2936289
Pages (from-to)	114349-114362
Number of pages	14
Journal	IEEE Access
Volume	7
DOIs	https://doi.org/10.1109/ACCESS.2019.2936289
Publication status	Published - 2019

Bibliographical note

Funding Information:
This work was supported in part by the Cross-Ministry Giga Korea Project Grant funded by the Korean Government (MSIT) (Development of 4D Reconstruction and Dynamic Deformable Action Model-Based Hyper-Realistic Service Technology) under Grant GK18P0200, and in part by the National Research Foundation of Korea (NRF) Grant funded by the Korean Government (MSIP) under Grant NRF-2018R1A2B3003896.

Publisher Copyright:
© 2020 Royal Society of Chemistry. All rights reserved.

Keywords

Aesthetic assessment
Convolutional neural network
Image aesthetics
Image composition
Unsupervised attribute clustering
Unsupervised property discovery

ASJC Scopus subject areas

General Computer Science
General Materials Science
General Engineering

Access to Document

10.1109/ACCESS.2019.2936289

Cite this

@article{a4be04c981344209a2707492ec1dc02c,

title = "Property-specific aesthetic assessment with unsupervised aesthetic property discovery",

abstract = "We propose the property-specific aesthetic assessment (PSAA) algorithm with unsupervised aesthetic property discovery. The proposed PSAA algorithm uses an aesthetic feature extractor, an aesthetic property classifier, and multiple property-specific assessment networks. The aesthetic feature extractor analyzes aesthetics of images to generate features. Using such aesthetic features, we discover diverse aesthetic properties in an unsupervised manner and develop the aesthetic property classifier to predict the aesthetic property of each image. For each discovered aesthetic property, we train a property-specific assessment network. Thus, we can assess the aesthetic quality of an image using the property-specific network that corresponds to its property. Experimental results on a large dataset show that the proposed PSAA algorithm achieves state-of-the-art aesthetic assessment performance. Furthermore, we demonstrate that PSAA is useful for improving aesthetic qualities of images in two applications: contrast enhancement and image cropping.",

keywords = "Aesthetic assessment, Convolutional neural network, Image aesthetics, Image composition, Unsupervised attribute clustering, Unsupervised property discovery",

author = "Lee, \{Jun Tae\} and Chul Lee and Kim, \{Chang Su\}",

note = "Funding Information: This work was supported in part by the Cross-Ministry Giga Korea Project Grant funded by the Korean Government (MSIT) (Development of 4D Reconstruction and Dynamic Deformable Action Model-Based Hyper-Realistic Service Technology) under Grant GK18P0200, and in part by the National Research Foundation of Korea (NRF) Grant funded by the Korean Government (MSIP) under Grant NRF-2018R1A2B3003896. Publisher Copyright: {\textcopyright} 2020 Royal Society of Chemistry. All rights reserved.",

year = "2019",

doi = "10.1109/ACCESS.2019.2936289",

language = "English",

volume = "7",

pages = "114349--114362",

journal = "IEEE Access",

issn = "2169-3536",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Property-specific aesthetic assessment with unsupervised aesthetic property discovery

AU - Lee, Jun Tae

AU - Lee, Chul

AU - Kim, Chang Su

N1 - Funding Information: This work was supported in part by the Cross-Ministry Giga Korea Project Grant funded by the Korean Government (MSIT) (Development of 4D Reconstruction and Dynamic Deformable Action Model-Based Hyper-Realistic Service Technology) under Grant GK18P0200, and in part by the National Research Foundation of Korea (NRF) Grant funded by the Korean Government (MSIP) under Grant NRF-2018R1A2B3003896. Publisher Copyright: © 2020 Royal Society of Chemistry. All rights reserved.

PY - 2019

Y1 - 2019

N2 - We propose the property-specific aesthetic assessment (PSAA) algorithm with unsupervised aesthetic property discovery. The proposed PSAA algorithm uses an aesthetic feature extractor, an aesthetic property classifier, and multiple property-specific assessment networks. The aesthetic feature extractor analyzes aesthetics of images to generate features. Using such aesthetic features, we discover diverse aesthetic properties in an unsupervised manner and develop the aesthetic property classifier to predict the aesthetic property of each image. For each discovered aesthetic property, we train a property-specific assessment network. Thus, we can assess the aesthetic quality of an image using the property-specific network that corresponds to its property. Experimental results on a large dataset show that the proposed PSAA algorithm achieves state-of-the-art aesthetic assessment performance. Furthermore, we demonstrate that PSAA is useful for improving aesthetic qualities of images in two applications: contrast enhancement and image cropping.

AB - We propose the property-specific aesthetic assessment (PSAA) algorithm with unsupervised aesthetic property discovery. The proposed PSAA algorithm uses an aesthetic feature extractor, an aesthetic property classifier, and multiple property-specific assessment networks. The aesthetic feature extractor analyzes aesthetics of images to generate features. Using such aesthetic features, we discover diverse aesthetic properties in an unsupervised manner and develop the aesthetic property classifier to predict the aesthetic property of each image. For each discovered aesthetic property, we train a property-specific assessment network. Thus, we can assess the aesthetic quality of an image using the property-specific network that corresponds to its property. Experimental results on a large dataset show that the proposed PSAA algorithm achieves state-of-the-art aesthetic assessment performance. Furthermore, we demonstrate that PSAA is useful for improving aesthetic qualities of images in two applications: contrast enhancement and image cropping.

KW - Aesthetic assessment

KW - Convolutional neural network

KW - Image aesthetics

KW - Image composition

KW - Unsupervised attribute clustering

KW - Unsupervised property discovery

UR - https://www.scopus.com/pages/publications/85086805449

U2 - 10.1109/ACCESS.2019.2936289

DO - 10.1109/ACCESS.2019.2936289

M3 - Article

AN - SCOPUS:85086805449

SN - 2169-3536

VL - 7

SP - 114349

EP - 114362

JO - IEEE Access

JF - IEEE Access

M1 - 2936289

ER -

Property-specific aesthetic assessment with unsupervised aesthetic property discovery

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this