The deep stack and tremendous amount of computational parameters in deep learning models greatly increases the challenges of pre-processing, training, testing, and post- processing geospatial datasets quickly and efficiently. This session will discuss the latest progresses on constructing advanced cyberinfrastructure for deep learning on satellite-based or field-observed geospatial datasets. The goal is to bring community experiences together and collaborate on building advanced geospatial cyberinfrastructure addressing the big questions raised in solving fundamental geoscience problems using deep learning models.
Presenter: Ziheng Sun
Presentation Title: Geoweaver for Better Deep Learning: A Review of Cyberinfrastructure
Slides: https://doi.org/10.6084/m9.figshare.9037091View Full Recording on YouTube