A survey and analysis on automatic image annotation

作者：

Highlights：

• A deep review on state-of-the-art methods in AIA.

• Discussions on open issues, main challenges, and promising directions.

• Detailed comparisons of the five types of AIA techniques.

摘要

•A deep review on state-of-the-art methods in AIA.•Discussions on open issues, main challenges, and promising directions.•Detailed comparisons of the five types of AIA techniques.

论文关键词：Automatic image annotation,Generative model,Nearest-neighbor model,Discriminative model,Tag-completion,Deep learning

论文评审过程：Received 19 April 2017, Revised 31 January 2018, Accepted 11 February 2018, Available online 13 February 2018, Version of Record 23 February 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.02.017