Design and Development of Advance Explainable AI Techniques for Change Detection from Remotely Sensed Images in Presence of Limited Training Samples
Implementing Organization
North-Eastern Hill University https://www.nehu.ac.in/
Principal Investigator
Dr. Anindya Halder
North-Eastern Hill University https://www.nehu.ac.in/
anindya.halder@gmail.com
CO-Principal Investigator
Prof. Mahesh Pal
North-Eastern Hill University https://www.nehu.ac.in/
Project Overview
The research develops explainable AI techniques for detecting building change and planetary surface change from remote sensing images with limited training data. It focuses on detecting natural environmental changes and man-made changes and extends the methods to planetary surfaces changes. The study uses semi-supervised learning, transformers, attention models, and generative approaches to analyze temporal satellite image data and predict future LULC changes. Explainable AI methods will ensure model transparency and support reliable environmental monitoring and planetary exploration.
Achievements
The project being undertaken by ISRO in collaboration with academia is directly or indirectly beneficial to the society.