RT Journal Article
A1 Xin Jin
A1 Le Wu
A1 Xiaodong Li
A1 Xiaokun Zhang
A1 Jingying Chi
A1 Siwei Peng
A1 Shiming Ge
A1 Geng Zhao
A1 Shuying Li

PB iet
T1 ILGNet: inception modules with connected local and global features for efficient image aesthetic quality classification using domain adaptation
JN IET Computer Vision
VO 13
IS 2
SP 206
OP 212
AB In this study, the authors address a challenging problem of aesthetic image classification, which is to label an input image as high- or low-aesthetic quality. We take both the local and global features of images into consideration. A novel deep convolutional neural network named ILGNet is proposed, which combines both the inception modules and a connected layer of both local and global features. The ILGnet is based on GoogLeNet. Thus, it is easy to use a pre-trained GoogLeNet for large-scale image classification problem and fine tune their connected layers on a large-scale database of aesthetic-related images: AVA, i.e. domain adaptation. The experiments reveal that their model achieves the state of the arts in AVA database. Both the training and testing speeds of their model are higher than those of the original GoogLeNet.
K1 low-aesthetic quality
K1 high-aesthetic quality
K1 connected global features
K1 ILGNet
K1 inception modules
K1 efficient image aesthetic quality classification
K1 large-scale image classification problem
K1 domain adaptation
K1 connected local features
K1 deep convolutional neural network
K1 AVA database
K1 large-scale aesthetic-related image database
K1 GoogLeNet
DO https://doi.org/10.1049/iet-cvi.2018.5249
UL https://digital-library.theiet.org/;jsessionid=1i4c9vh9690vl.x-iet-live-01content/journals/10.1049/iet-cvi.2018.5249
LA English
SN 1751-9632
YR 2019
OL EN