Deep Learning for the Earth Sciences
© 2021 Wayra
Deep learning methods have only recently made their way into the earth science programs taught at universities. A new book titled Deep Learning for the Earth Sciences provides a comprehensive overview of how these methods can be applied to the earth sciences, through contributions by a broad panel of international experts. EPFL professor Devis Tuia is one of the book’s four editors.
Deep learning is a fundamental technique in modern artificial intelligence and is being applied to disciplines across the scientific spectrum; the earth sciences are no exception. Deep Learning for the Earth Sciences provides a pioneering, unifying perspective to the application of deep learning methods because it’s the first to bring together insights from the world’s leading experts on this issue. The book is intended for informed readers and will enable PhD students and researchers to quickly become familiar with modern approaches. Devis Tuia, an associate professor at EPFL’s Environmental Computational Science and Earth Observation Laboratory, is one of the four editors of the book, along with Gustau Camps-Valls from Universitat de València, Spain, Xiaoxiang Zhu from the German Aerospace Center and Technical University of Munich, Germany, and Markus Reichstein from the Max Planck Institute, Germany.
An international initiative pulling together an entire community
Deep Learning for the Earth Sciences was made possible thanks to the concerted efforts of a large number of experts around the world. “We made sure that the authors reviewed each other’s contributions so that our book isn’t just a collection of articles on high-tech subjects, but a coherent reference that gives readers a solid understanding of the current state of deep learning methods,” says Tuia, who also sits on the steering committee at the EPFL Center for Imaging. “We found that much of the technology used today – such as convolutional networks, unsupervised pre-training and recurrent networks – cuts across the disciplines explored in the book. All the experts involved have appropriated this technology, each in their own way, so as to apply it to their field of research. This means going beyond technology transfer and adapting deep-learning methods to the physical constraints of the problem they’re working on, for instance, or studying the causal relationships that make these methods more efficient and transparent. This gives earth scientists a better understanding of our environment and enables them to better explain the processes they observe in images.”
The book is available both online and in print form, and provides new ideas and recommendations for further research.