FIS-GAN: GAN with Flow-based Importance Sampling
This work addresses a computational bottleneck in GAN training for researchers and practitioners, though it is incremental as it adapts existing importance sampling techniques to adversarial learning.
The paper tackles the inefficiency of GAN training by replacing uniform or Gaussian sampling in the latent space with importance sampling, using normalizing flow for density estimation, and demonstrates that this method significantly accelerates optimization while maintaining visual fidelity on MNIST and Fashion-MNIST datasets.
Generative Adversarial Networks (GAN) training process, in most cases, apply Uniform or Gaussian sampling methods in the latent space, which probably spends most of the computation on examples that can be properly handled and easy to generate. Theoretically, importance sampling speeds up stochastic optimization in supervised learning by prioritizing training examples. In this paper, we explore the possibility of adapting importance sampling into adversarial learning. We use importance sampling to replace Uniform and Gaussian sampling methods in the latent space and employ normalizing flow to approximate latent space posterior distribution by density estimation. Empirically, results on MNIST and Fashion-MNIST demonstrate that our method significantly accelerates GAN's optimization while retaining visual fidelity in generated samples.