Image Captioning Model Using Part-of-Speech Guidance Module for Description With Diverse Vocabulary

Image captions aim to generate human-like sentences that describe the image’s content. Recent developments in deep learning (DL) have made it possible caption images for accurate descriptions and detailed expressions. However, since DL learns relationship between captions, constructs based on commonly frequented words dataset. Although these generated are highly accurate, they low lexical diversity, unlike humans due limited vocabulary. Therefore, this paper, we propose a Part-Of-Speech (POS) guidance module multimodal-based image captioning model determines intensity of word sequences generates through POS enhance diversity DL. The proposed enables rich expression by controlling information predicted predict words. Then, multimodal layer adds output vector Bi-LSTM using next caption, considering grammatical structure. We trained tested Flicker 30K MS COCO datasets compared them with current state-of-the-art studies. Also, analyzed Type-Token Ratio (TTR) confirmed several