Spoofing Attack Augmentation: Can Differently-Trained Attack Models Improve Generalisation? | IEEE Conference Publication | IEEE Xplore