TY - CONF AU - Effenberger, Frederic AU - Vasile, Ruggero AU - Cherti, Mehdi AU - Kesselheim, Stefan AU - Jitsev, Jenia TI - Generative Adversarial Deep Learning with Solar Images M1 - FZJ-2022-00868 SP - 1 PY - 2021 AB - The Solar Dynamics Observatory (SDO) offers an unprecedented, very large dataset of solar images in different optical and EUV wavelength bands, capturing solar atmospheric structures in high resolution and with excellent coverage since 2010. This dataset is thus well suited to study the application of advanced machine learning techniques that require large amounts of data for training, such as deep learning approaches. Here, we present results of generative adversarial deep learning as applied to a large database of solar images and discuss challenges in training and validation, in particular with distributed training on large computer clusters. We address the potential of data augmentation techniques for improved learning and image quality and the opportunities for latent space structure exploration and control. Potential application downstream that can make use of such generated images are briefly discussed and the need for a community-driven, physics-based basis to establish evaluation criteria for generative models will be emphasized. T2 - AGU Fall Meeting CY - 13 Dec 2021 - 17 Dec 2021, New Orleans (USA) Y2 - 13 Dec 2021 - 17 Dec 2021 M2 - New Orleans, USA LB - PUB:(DE-HGF)8 UR - https://juser.fz-juelich.de/record/905641 ER -