The School of Computer Science is pleased to present…
Date: Thursday, March 13th, 2025
Time: 1:00 PM
Location: Chrysler Hall South, Room 51
Monitoring and comparing coastal environments remain challenging due to the lack of domain-specific datasets that capture their unique features. General-purpose models, including scene graph generation techniques, struggle to extract meaningful semantic details from these landscapes. This research addresses this gap by developing a domain-specific dataset tailored to coastal environments and proposing a pipeline for generating coastal similarity scores. By leveraging scene graph generation, this work enables more meaningful comparisons beyond superficial visual features. The resulting framework aims to enhance environmental monitoring efforts and contribute to the broader field of AI-driven landscape analysis.
Internal Reader: Dr. Boubakeur Boufama
External Reader: Dr. Mohammad Hassanzadeh
Advisors: Dr. Ziad Kobti & Dr. Chris Houser
