Kerai, Shivani
Unknown Affiliation

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Contextual embedding generation of underwater images using deep learning techniques Kerai, Shivani; Khekare, Ganesh
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 13, No 3: September 2024
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v13.i3.pp3111-3118

Abstract

This article delves into the cutting-edge realm of artificial intelligence, specifically focusing on its application in marine research via underwater image analysis. It introduces an innovative, integrated approach that combines object detection with image captioning tailored for the aquatic domain. Central to this approach is the advanced technique of image feature extraction, complemented by the strategic implementation of attention mechanisms within neural networks. These mechanisms are key in enhancing the precision and contextual understanding of underwater imagery. The efficacy of this method is underscored by extensive experiments on diverse underwater datasets. Results show notable improvements in detecting and describing complex underwater scenes, thereby providing invaluable insights for marine biologists, environmentalists, and the broader scientific community. This exploration marks a significant advancement in marine research, offering a new lens through which the underwater world can be understood and preserved.