Deep Learning in Astronomy

This project represents a collaboration with the University of Sciences to address two critical challenges in modern astronomy through Deep Learning techniques.

The main objective is to develop artificial intelligence models capable of analyzing images of the most distant galaxies captured by the Hubble Space Telescope, providing accurate information about their morphology and detecting their edges with unprecedented precision.

Challenges Addressed

• Edge Detection: Precise identification of galaxy boundaries in high-resolution Hubble images
• Morphological Classification: Automatic analysis of the structure and shape of distant galaxies
• Large Volume Processing: Efficient analysis of extensive catalogs of astronomical images

Technologies Used

The project uses PyTorch to implement deep neural networks specialized in astronomical image processing, combining computer vision techniques with astronomical knowledge to achieve state-of-the-art results.

The Hubble Space Telescope (HST) is one of the most important scientific instruments ever built, launched into space in 1990 by NASA and ESA. Orbiting at approximately 547 kilometers above Earth, the Hubble has revolutionized our understanding of the universe by capturing images of extraordinary clarity, free from the atmospheric distortion that affects ground-based telescopes.