Classification of hyperspectrales 3D data with machine-learning

Hyper-spectral imaging is highly sought after in many fields including geology, physics, astronomy and biomedical imaging. Hence, the volume of of hyper-spectral data has skyrocketed over the past decade in all fields. This surge in large volumes of data related to 3D data calls for efficient classification methods in order to explore these large hyper-spectral data-sets without a priori assumptions and classification is still undoubtedly the bottleneck in many areas. Thus scientists in all fields are going to be naturally dependent of advancements in modern machine learning techniques. Given the wide range of applications, it is thus urgent to develop versatile algorithms to classify the 3D spectra with ML techniques (such as CNN for supervised or self-supervised classifications). We aim to bring experts in computing ML, and users of 3D hyperspectral data in physics and astrophysics to share expertise around the techniques of 3D classification.