Presentation #201.13 in the session Star Clusters and Associations — iPoster Session.
Galaxies are the smallest building blocks of the large scale structure of the Universe. They present as extended objects in observational images as a result of which they can be classified into different shape-types or morphologies. Galaxy morphologies correlate strongly with various dynamical properties of galaxies as well as potentially provide markers for large scale structure and evolution of Universe. In the poster, we present the results for the classification of galaxies between spirals, ellipticals, and irregulars using a supervised Deep learning model based on the Vision Transformer architecture. In this novel approach, we use a pre-labelled dataset for training and predict classifications on a set of test data. We find that the Vision Transformer based Deep learnt model performs a very good and accurate classification of the different morphological types of galaxies as indicated by various performance metrics.