Jocelyn Chanussot – författare

Advances in Machine Learning and Image Analysis for GeoAI presents recent advances in applications and algorithms that are at the intersection of Geospatial imaging and Artificial Intelligence (GeoAI). The book covers algorithmic advances in geospatial image analysis, sensor fusion across modalities, few-shot open-set recognition, explainable AI for Earth Observations, self-supervised learning, image superresolution, Visual Question Answering, and spectral unmixing, among other topics. This book offers a comprehensive resource for graduate students, researchers, and practitioners in the area of geospatial image analysis. It provides detailed descriptions of the latest techniques, best practices, and insights essential for implementing deep learning strategies in GeoAI research and applications.

Covers the latest machine learning and signal processing techniques that can effectively leverage multimodal geospatial imagery at scaleChapters cover a variety of algorithmic frameworks pertaining to GeoAI, including superresolution, self-supervised learning, data fusion, explainable AI, among othersPresents cutting-edge deep learning architectures optimized for a wide array of geospatial imagery

The role of artificial intelligence is crucial in the domain of Earth Observation (EO) data analysis. Deep learning-based approaches have improved accuracy, but they have affected the reliability and transparency of EO data. It is critical to improve the explainability of EO data analysis algorithms and complex deep learning models to ensure the quality of spatial decisions. This book discusses the various advancements in Explainable AI and investigates their suitability for various EO data analyses offering best practices for implementing algorithms that facilitate big and efficient data processing. It lays the foundation of Explainable EO and helps readers build trustworthy, secure, and robust EO systems.Features:Examines explainability of algorithms from the aspect of generalizability and reliabilityReviews state-of-the-art explainability strategies related to the preprocessing algorithmsProvides explanations for specific evaluation metrics of various EO data processing and preprocessing algorithmsDiscusses explainable ante-hoc and post-hoc approaches for EO data analysisServes as a foundational reference for developing future EO data processing strategiesAddresses the key challenges in making EO data processing algorithms interpretable and offers insights for the future of explainable EO data processingThis book is intended for graduate students, researchers and academics in computer and data science, machine learning, and image processing, as well as professionals in geospatial data science using GIS and remote sensing in Earth and environmental sciences.

Multivariate imagery is now a very common tool in numerous applications, ranging from satellite remote sensing and astrophysics to biomedical imagery, monitoring of the environment or industrial inspection. Multivariate must be understood in th emost general way: color and multispectral imaging, but also multimodal, multisource or multitemporal imagery. In all the cases, the multivariate image corresponds to a set of standard grey level images. The availability of the additional diversity, be it spectral temporal and s.o., provides an invaluable source of information, enabling to consider a wide range of new applications. However, in order to address these applications, theoretical developments are required in terms of signal and image processing, or, more generally speaking, information processing. As a matter of fact, most of the standard algorithms designed for grey level images do not generalize easily to multidimensional spaces and some specific derivations are required.This book aims at presenting the most recent advances in signal and image processing for the analysis of multivariate data. It should be helpful for electrical engineers, PhD students and researcher working in the field of signal processing, but also for any engineer dealing with some specific application where multidimensional data are processed.

This book reviews the state of the art in algorithmic approaches addressing the practical challenges that arise with hyperspectral image analysis tasks, with a focus on emerging trends in machine learning and image processing/understanding.

This book reviews the state of the art in algorithmic approaches addressing the practical challenges that arise with hyperspectral image analysis tasks, with a focus on emerging trends in machine learning and image processing/understanding.

This book contains the thoroughly refereed proceedings of the 12th International Symposium on Mathematical Morphology, ISMM 2015 held in Reykjavik, Iceland, in May 2015. The 62 revised full papers were carefully reviewed and selected from 72 submissions. The papers are organized in topical sections on evaluations and applications;

Optical remote sensing relies on exploiting multispectral and hyper spectral imagery possessing high spatial and spectral resolutions respectively. These modalities, although useful for most remote sensing tasks, often present challenges that must be addressed for their effective exploitation. This book presents current state-of-the-art algorithms that address the following key challenges encountered in representation and analysis of such optical remotely sensed data.Challenges in pre-processing images, storing and representing high dimensional data, fusing different sensor modalities, pattern classification and target recognition, visualization of high dimensional imagery.