Skip to main content
. Author manuscript; available in PMC: 2022 Aug 14.
Published in final edited form as: KDD. 2021 Aug 14;2021:617–627. doi: 10.1145/3447548.3467281

Figure 2:

Figure 2:

Comparison of vanilla adversarial loss versus the squared adversarial loss on MNIST-to-USPS (top) and USPS-to-MNIST (bottom) UDA. We vary the probability of target users. For both UDA experiments, the SOTA central methods [23] can achieve over 98% accuracies. From left to right, the columns are target domain accuracies, classification losses and adversarial losses of target domain users.