Softmax GAN
This addresses training instability for GAN users, though it appears incremental as a variant of existing GAN methods.
The paper tackles the instability in Generative Adversarial Network (GAN) training by introducing Softmax GAN, which replaces the classification loss with a softmax cross-entropy loss in batch sample space, resulting in stabilized training as demonstrated experimentally.
Softmax GAN is a novel variant of Generative Adversarial Network (GAN). The key idea of Softmax GAN is to replace the classification loss in the original GAN with a softmax cross-entropy loss in the sample space of one single batch. In the adversarial learning of $N$ real training samples and $M$ generated samples, the target of discriminator training is to distribute all the probability mass to the real samples, each with probability $\frac{1}{M}$, and distribute zero probability to generated data. In the generator training phase, the target is to assign equal probability to all data points in the batch, each with probability $\frac{1}{M+N}$. While the original GAN is closely related to Noise Contrastive Estimation (NCE), we show that Softmax GAN is the Importance Sampling version of GAN. We futher demonstrate with experiments that this simple change stabilizes GAN training.