Aesthetic evaluation of Asian and Caucasian photos with overall and attribute scores. (October 2022)

• Record Type: Journal Article
• Authors: Huang, Heng; Jin, Xin; Li, Xinning; Cui, Shuai; Xiao, Chaoen
• Is Part Of: Computers & electrical engineering. Volume 103(2022)
• Publication Date: 2022-10
• Subjects: Image aesthetic -- Portrait -- Aesthetic attribute -- Neural network; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.108341
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: Nowadays, the portrait image aesthetic attribute evaluation is extremely valuable in the research area, but there is no either specialized portrait image aesthetic attribute dataset or specialized portrait image aesthetic dataset. At the same time, the lightweight image aesthetic attribute evaluation model is also valuable in the market. In this paper, we propose two datasets about portrait image aesthetic attribute scores and portrait image overall aesthetic scores. We also design two models to predict attribute scores and overall scores. We use Knowledge Distillation(KD) and train our feature extraction model on a large-scale dataset to make the evaluation models lightweight. In the experiments, our models have good performance in experiments and extremely highly practical value. The parameter quantity of our models is all less than 6 M and the Spearman's Rank Correlation Coefficient (SRCC) of results are above 0.75. These models provide aesthetic references to portrait images, which can give some photographic suggestions for color, lighting, and composition.