jakgel

Semantic Segmentation

• Progressively Growing Generative AdversarialNetworks for High Resolution Semantic Segmentation of Satellite Images

Map Generation

• GeoGAN: A Conditional GAN with Reconstructionand Style Loss to Generate Standard Layer of Maps from Satellite Images

jakgel

Machine learning and deep learning in astronomy: from 2016, sparse after after mid 2019

Machine learning has a long history in astronomy, but since around 2016 deep learning (DL) only got traction. Here I summarise effords and publications of DL in the astrophysical community.

The main field of ML applications in astrophysics is object classification. With source counts now ranging into the 10⁷-10⁸ for most surveys, machine learning is put to use to allow the classification of a large number of sources which would otherwise need an infeasible amount of manpower:

• CATALOGING ACCRETED STARS WITHIN GAIA DR2 USING DEEP LEARNING 07/2019 - Three hidden-layer FCN with 4-9 initial input features, trained on simulations for 2-class classification.
• Classifying galaxies according to their HI content 06/2019 - SVM perform best in specisifity and is hence used over a DNN
• [Morphological classification