Speaker
Description
Identifying the extent of internal galaxy structures can be incredibly helpful in unravelling their evolutionary histories. For bars in particular this can allow for precise measurements of colour, stellar mass, bar length and alignment with the disk. We present ZooBot:3D, a deep learning model trained using the volunteer classifications from the Galaxy Zoo: 3D project, capable of generating highly detailed segmentation maps of disk galaxies. We employ this model to produce maps for ~500,000 galaxies in the DESI Legacy Survey and demonstrate how these masks can be used to study the intricate properties of barred and unbarred spirals. Finally, we will detail the forthcoming public data release of these segmentation products and lay out plans for future work in this area.
