A Review of Stylized Image Captioning Techniques, Evaluation Parameters, and Datasets

Image captioning is generation of natural language sentences for a given image. With recent success of deep-learning models for image captioning, many researchers have contributed towards Stylized Image Captioning (SIC). SIC is an emerging topic in which system generates a natural language description of an image that incorporates a specific style (e.g., humor, romance, fear, surprise, etc.). In this paper, a precise review on Stylized Image Captioning is presented. This review focusses on various challenges to SIC and different existing methods. Further, different evaluation parameters with benchmark datasets for SIC are discussed. Finally, the paper explores future directions in the area of SIC. Further, researchers can use this review paper as a guide to stay current on the most recent developments in the field of SIC.