Visual Paragraph Generation: Review

Abstract

Visual Paragraph generation is the field of generating human-like detailed paragraph, consisting of multiple sentences rather than one sentence, that fully describes the visual content. In line with its growing applications in various domains, there has been an increase in the number of research works conducted to improve the quality of the generated description for a given visual content. Even though good results were achieved, there are still some challenges to be overcome especially with the dawn of big data and its influence on existing information systems. This paper is to create familiarity with the current thinking and research in visual paragraph generation by providing a thorough survey to the literature. Its main aims are to highlight recent contributions and achievements, identify the promising technologies that support future research, and justify the need for further research work in this field.