Text this: Leveraging vision transformer architectures for glacial and fluvio-glacial landform classification.