Generating Semantically Correct Hindi Captions Using Deep Neural Network

Singh, Akash Ramsingh (2021) Generating Semantically Correct Hindi Captions Using Deep Neural Network. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Image captioning is one of the most significant and exciting challenges in computer vision and natural language processing. Several studies have been conducted in this field, the majority of which have focused on the English language. Foreign language research has a wide range of possibilities. This image captioning research is being conducted for the language Hindi. The research makes use of the Flickr-8k dataset’s machine-translated Hindi captions. The research is carried out using an encoder-decoder framework. Image features are extracted using pre-trained CNNs such as VGG16, ResNet50, and Inception V3. Uni-directional and Bi-directional LSTM is employed for the text encoding process. A thorough comparison is made between various LSTM and Bi-LSTM models in this research. The VGG16 with the bi-LSTM model performed the best by giving a BLUE1 score of 0.583.

Item Type:	Thesis (Masters)
Subjects:	P Language and Literature > PK Indo-Iranian Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	11 Mar 2023 13:39
Last Modified:	11 Mar 2023 13:39
URI:	https://norma.ncirl.ie/id/eprint/6311

Actions (login required)

View Item