CaDS Seminar 2024 - Jul. 2
Prof. Dr. Gabriele Cavallaro (SDL AI and ML for Remote Sensing)
Advancing Geoscience through Large-Scale Machine Learning and Remote Sensing with Supercomputing: Challenges and Opportunities
Abstract:
The rapid proliferation of data in the new information era has increased the complexity of data-driven problems across various fields of science and engineering. This development has led to a paradigm shift in Machine Learning (ML), moving towards unsupervised and self-supervised representation learning, as well as multimodal learning. Significant advancements have emerged not only in mainstream Natural Language Processing and Computer Vision but also in Earth observation (EO) applications. These advancements exploit the synergies between self-supervised learning and the expanded availability of supercomputing systems, resulting in the emergence of Foundation Models (FMs). Originating from the concept of building upon an existing “foundation,” these models are developed by training on large and diverse datasets. This training enables them to capture a broad spectrum of informative features, making them extremely versatile and potentially applicable across multiple domains. This presentation provides an overview of the efforts and activities of the SDL “AI and ML for Remote Sensing,” currently involved in several international projects aimed at developing the next generation of geospatial FMs for EO. It will highlight the challenges and opportunities in conducting interdisciplinary research that intersects ML, supercomputing, and remote sensing applications.