I-GUIDE VCO: Geospatial Transfer Learning on I-GUIDE Platform

Geospatial Transfer Learning on I-GUIDE Platform

June 6, 2023 12:00 pm (Central Time)


Recorded VCO


Abstract

This VCO session demonstrates the I-GUIDE platform for applying transfer learning to solve the problem of hydrographic streamline delineation. By utilizing the segmentation models library within the I-GUIDE platform, we showcase how pre-trained models from ImageNet classification can be effectively adapted to enhance the efficiency and accuracy of streamline delineation tasks. The session discusses how to utilize the I-GUIDE platform for transfer learning applications while providing researchers and practitioners with an accessible solution for image segmentation workflows. Attendees will gain insights into a practical implementation of geospatial transfer learning techniques that could be applied to various application domains.

Speakers

Nattapon Jaroenchai

Nattapon Jaroenchai

University of Illinois Urbana-Champaign

Nattapon Jaroenchai is a Ph.D. student in the Geography and Geographic Information Science department, advised by Dr. Shaowen Wang at the University of Illinois Urbana-Champaign (UIUC). His research develops geospatial AI and machine learning approaches, mainly transfer learning techniques, to solve complex environmental problems. Nattapon’s recent work focuses on resolving geospatial data integration and uncertainty challenges in deep learning models to advance hydrological and topographic knowledge discovery. He holds a Master of Science in Geography and Geographic Information Science from UIUC and a Bachelor’s in Computer Engineering from Chulalongkorn University.