ML for HyperSpectral Imaging (ML4HSI) Project
Research on deep learning applications in 3D astrophysical data analysis.
Project Overview
The ML4HSI project explores the use of deep learning techniques for analyzing 3D spectroscopic data in astrophysics. This includes developing neural networks for tasks such as galaxy morphology classification, kinematic modeling, and anomaly detection in large-scale surveys.
Key aspects of the research involve:
- Convolutional Neural Networks (CNNs) adapted for 3D data cubes
- Integration with existing tools and methods like Non-Negative Matrix Factorization
- Applications to MUSE data
Related Links
For more information, visit the Artificial Inteligence for Spectral and Hyperspectral Imaging (AISHI) project website and the 2024 Deep Learning 3D conference website.
Publications:- Bourahma et al. 2026, A&A, submitted "Accurate spectroscopic redshift estimation using non-negative matrix factorization: application to MUSE spectra"