This book is a comprehensive guide that bridges the gap between foundational principles and cutting-edge advancements in hyperspectral imaging and deep learning. With contributions from leading international experts, this book covers a wide range of topics essential for researchers, engineers, and professionals in the field.

This book begins with an introduction to hyperspectral imaging and deep learning, setting the stage for more advanced discussions. Subsequent chapters delve into neural network architectures, graph-based methods, generative models, and the application of transformers in hyperspectral imaging. Each chapter not only presents theoretical insights but also practical applications, making complex concepts accessible and relevant.

Readers will discover methods to optimize deep learning models through techniques like quantization and pruning, ensuring efficiency without sacrificing performance. Additionally, this book addresses the practical challenges of managing and processing large volumes of hyperspectral data, offering strategies for data storage, management, and parallel processing.

Exclusive online resources, including example codes, tutorials, and hyperspectral datasets, complement the comprehensive content, enabling readers to apply what they learn in real-world scenarios. This book is an indispensable resource for anyone looking to harness the power of hyperspectral technology to drive innovation and solve complex problems.

Is a comprehensive guide from theoretical basics to advanced applications Provides access to example codes, tutorials, and hyperspectral datasets for immediate hands-on practice Offers a global and authoritative perspective in hyperspectral image analysis

Autorentext

Juan Mario Haut is an associate professor in the Department of Computer and Communication Technology at the University of Extremadura. The topic of his research covers the efficient analysis of remotely sensed (RS) images collected from Earth’s surface Observation platforms through the design and implementation of novel machine (ML) and deep learning (DL) processing methods. Dr. Haut delves into the application of Big Data and High-performance Computing (HPC) strategies, such as parallelization and distribution over GPU devices and Cloud Computing platforms, combined with deep neural networks for large and complex RS dataset analysis, such as hyperspectral and multispectral images. Dr. Haut is an author and a co-author of more than 120 scientific publications, including more than 70 contributions to JCR journals and more than 50 contributions to congresses, both national (17) and international (36) of relevance such as IEEE IGARSS, IEEE WHISPERS, or IEEE CBMS, and 1 book chapter. .

M.E. Paoletti is a professor in the Department of Computer and Communication Technology at the University Centre of Merida, University of Extremadura, and a researcher at the Hyperspectral Computing Laboratory (HyperComp). Her research focuses on the efficient processing of remote sensed hyperspectral images through the development of deep learning techniques combined with graphical processing. Dr. Paoletti is author and co-author of 117 scientific publications, including 70 contributions to JCR journals, 49 contributions to congresses, both national (16) and international (33) of relevance such as IEEE IGARSS, IEEE WHISPERS, or IEEE CBMS, and 1 book chapter. Her JCR contributions stand out in the fields of computation (e.g., Journal of Supercomputing), neural networks (e.g., IEEE TNNLS and Neurocomputing), and remote sensing (e.g., IEEE TGRS, IEEE GRSL, or IEEE GRSM), having 8 highly cited articles with 2 research fronts (InCites Essential Science Indicators of Clarivate).

Inhalt

Introduction to Hyperspectral Imaging and Deep Learning.- Advances in Deep Neural Architectures for Hyperspectral Image Analysis.- Graph-Based Methods for Hyperspectral Data Analysis.- Generative Models for Hyperspectral Imaging Processing.- Transformers and Foundation Models for Hyperspectral Imaging.- Efficient Deep Learning for Hyperspectral Image Classification: Quantization, Distillation, and Pruning.- Handling Large Volumes of Hyperspectral Data.