An Integrative Review of Image Captioning Research. Issue 4 (January 2021)

• Record Type: Journal Article
• Authors: Wang, Chaoyang; Zhou, Ziwei; Xu, Liang
• Is Part Of: Journal of physics. Volume 1748:Issue 4(2021)
• Publication Date: 2021-01
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1748/4/042060
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: In the field of computer vision, image captioning is a new frontier research. The basic task of image captioning is to generate a descriptive natural language for the input image. This paper investigates and analyzes the related research of image captioning. Firstly, the task and application scenarios of image captioning are introduced. Secondly, the image captioning algorithm based on template and the image captioning algorithm based on encoder-decoder structure are analyzed, and the advantages and limitations of each method are discussed. Then, the benchmark dataset and evaluation for image captioning are introduced. Finally, the future development of image captioning is prospected.