Modeling content-attribute preference for personalized image esthetics assessment

作者：

Highlights：

• A new personalized image esthetics assessment model is proposed.

• Esthetic preference is modeled using personality and content-attribute features.

• Content, attribute and personality tasks are pre-trained for feature extraction.

• The proposed method achieves better performance than the state-of-the-arts.

摘要

•A new personalized image esthetics assessment model is proposed.•Esthetic preference is modeled using personality and content-attribute features.•Content, attribute and personality tasks are pre-trained for feature extraction.•The proposed method achieves better performance than the state-of-the-arts.

论文关键词：Image esthetics assessment,Esthetic preference,Personality

论文评审过程：Received 26 April 2022, Revised 25 May 2022, Accepted 6 June 2022, Available online 9 June 2022, Version of Record 17 June 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2022.104505