Image Aesthetic Assessment Based on Latent Semantic Features

Yan, Gang; Bi, Rongjia; Guo, Yingchun; Wei-feng, Peng

doi:10.3390/info11040223

Cited by 3 publications

(3 citation statements)

References 37 publications

(67 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They proposed ILG-NET, combing the InceptionNet and the connected layer. Yan et al [24] obtained the aesthetic image features, including semantics, texture, and color. They weighted points of interest (POI) and segmented the image pixels.…”

Section: Related Workmentioning

confidence: 99%

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

Zhang¹,

Jiang²,

Song³

et al. 2023

Preprint

View full text Add to dashboard Cite

Image aesthetic assessment (IAA) with neural attention has made significant progress due to its effectiveness in object recognition. Current studies have shown that the features learned by convolutional neural networks (CNN) at different learning stages indicate meaningful information. The shallow feature contains the low-level information of images and the deep feature perceives the image semantics and themes. Inspired by this, we propose a visual enhancement network with feature fusion (FF-VEN). It consists of two sub-modules, the visual enhancement module (VE module) and the shallow and deep feature fusion module (SDFF module). The former uses an adaptive filter in the spatial domain to simulate human eyes according to the region of interest (ROI) extracted by neural feedback. The latter not only takes out the shallow feature and the deep feature by transverse connection, but also uses a feature fusion unit (FFU) to fuse the pooled features together with the aim of information contribution maximization. Experiments on standard AVA dataset and Photo.net dataset show the effectiveness of FF-VEN.

show abstract

Section: Related Workmentioning

confidence: 99%

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

Zhang¹,

Jiang²,

Song³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…They proposed ILG-NET, combing the InceptionNet and the connected layer. Yan et al [26] obtained the aesthetic image features, including semantics, texture, and color. They weighted points of interest (POI) and segmented the image pixels.…”

Section: Related Workmentioning

confidence: 99%

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

et al. 2023

View full text Add to dashboard Cite

Image aesthetic assessment (IAA) with neural attention has made significant progress due to its effectiveness in object recognition. Current studies have shown that the features learned by convolutional neural networks (CNN) at different learning stages indicate meaningful information. The shallow feature contains the low-level information of images, and the deep feature perceives the image semantics and themes. Inspired by this, we propose a visual enhancement network with feature fusion (FF-VEN). It consists of two sub-modules, the visual enhancement module (VE module) and the shallow and deep feature fusion module (SDFF module). The former uses an adaptive filter in the spatial domain to simulate human eyes according to the region of interest (ROI) extracted by neural feedback. The latter not only extracts the shallow feature and the deep feature via transverse connection, but also uses a feature fusion unit (FFU) to fuse the pooled features together with the aim of information contribution maximization. Experiments on standard AVA dataset and Photo.net dataset show the effectiveness of FF-VEN.

show abstract

“…The traditional product modeling design process is mainly designed from the product function to improve the product appearance form, including the design aspects of product form design, product color design, and product modeling texture [5][6]. The designer needs first to take the user's needs as the design direction, analyze the principle and performance of the product using their own design experience, and design the basic structure, function, and form of the corresponding product and other styling design elements, relying mainly on the designer's ability [7][8][9]. It is difficult to ensure the efficiency of design work and meet the requirements of rapid development and design product modeling design when done by designers alone.…”

Section: Introductionmentioning

confidence: 99%

Study on the aesthetic evaluation method of applying symmetry analysis in industrial product styling design

Wen,

Bai

2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

The evaluation of product styling design is transformed into a complex nonlinear solution problem by symmetry analysis in this paper. The equilibrium matrix decomposition evaluation system calculates the stored matrix singular values in product styling design and proves that symmetry can be obtained from the stress matrix row vector. The regression algorithm is used to process multiple industrial product design objectives, and each design element in product styling is coded to obtain the fitness evaluation of individual industrial products. It is found that the satisfaction score in product styling design is 97.4, the weight index of logo novelty element is up to 7, and the regression coefficient of plane shape is 0.269, to improve the scientific and feasibility of aesthetic evaluation of industrial product styling design.

show abstract

Image Aesthetic Assessment Based on Latent Semantic Features

Cited by 3 publications

References 37 publications

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

A Visual Enhancement Network with Feature Fusion for Image Aesthetic Assessment

Study on the aesthetic evaluation method of applying symmetry analysis in industrial product styling design

Contact Info

Product

Resources

About