A pretrained ntsnet
net = netsnet(pretrained=True, **{'topN': 6, 'device':'cpu', 'num_classes': 200})
You can read the full paper at this link.
@INPROCEEDINGS{Gallo:2019:IVCNZ,
author={Nawaz, Shah and Calefati, Alessandro and Caraffini, Moreno and Landro, Nicola and Gallo, Ignazio},
booktitle={2019 International Conference on Image and Vision Computing New Zealand (IVCNZ 2019)},
title={Are These Birds Similar: Learning Branched Networks for Fine-grained Representations},
year={2019},
month={Dec},
}