BOLT-GAN: Bayes-Optimal Loss for Stable GAN Training
This addresses training stability issues in generative adversarial networks for image generation, though it appears incremental as a modification of the existing WGAN framework.
The paper tackled the problem of unstable GAN training by introducing BOLT-GAN, a modification of WGAN inspired by the Bayes Optimal Learning Threshold, which achieved 10-60% lower Frechet Inception Distance on standard image generation benchmarks.
We introduce BOLT-GAN, a simple yet effective modification of the WGAN framework inspired by the Bayes Optimal Learning Threshold (BOLT). We show that with a Lipschitz continuous discriminator, BOLT-GAN implicitly minimizes a different metric distance than the Earth Mover (Wasserstein) distance and achieves better training stability. Empirical evaluations on four standard image generation benchmarks (CIFAR-10, CelebA-64, LSUN Bedroom-64, and LSUN Church-64) show that BOLT-GAN consistently outperforms WGAN, achieving 10-60% lower Frechet Inception Distance (FID). Our results suggest that BOLT is a broadly applicable principle for enhancing GAN training.