Name
Abstract | ||
|---|---|---|
Recent work has proposed various adversarial losses for training generative adversarial networks. Yet, it remains unclear what certain types of functions are valid adversarial loss functions, and how these loss functions perform against one another. In this paper, we aim to gain a deeper understanding of adversarial losses by decoupling the effects of their component functions and regularization terms. We first derive some necessary and sufficient conditions of the component functions such that the adversarial loss is a divergence-like measure between the data and the model distributions. In order to systematically compare different adversarial losses, we then propose DANTest, a new, simple framework based on discriminative adversarial networks. With this framework, we evaluate an extensive set of adversarial losses by combining different component functions and regularization approaches. This study leads to some new insights into the adversarial losses. For reproducibility, all source code is available at this https URL . |
Year | Venue | DocType |
|---|---|---|
2019 | arXiv: Learning | Journal |
Volume | Citations | PageRank |
abs/1901.08753 | 1 | 0.35 |
References | Authors | |
21 | 2 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Hao-Wen Dong | 1 | 20 | 3.50 |
| Yi-Hsuan Yang | 2 | 1022 | 84.71 |