Generator Knows What Discriminator Should Learn in Unconditional GANs
This work addresses the problem of enhancing fidelity in unconditional image generation for AI researchers, though it is incremental as it builds on existing GAN frameworks with a novel regularization approach.
The paper tackles the challenge of improving unconditional image generation by proposing a generator-guided discriminator regularization (GGDR) method, which uses generator feature maps as dense supervision for the discriminator, resulting in consistent performance improvements across multiple datasets in terms of quantitative and qualitative metrics.
Recent methods for conditional image generation benefit from dense supervision such as segmentation label maps to achieve high-fidelity. However, it is rarely explored to employ dense supervision for unconditional image generation. Here we explore the efficacy of dense supervision in unconditional generation and find generator feature maps can be an alternative of cost-expensive semantic label maps. From our empirical evidences, we propose a new generator-guided discriminator regularization(GGDR) in which the generator feature maps supervise the discriminator to have rich semantic representations in unconditional generation. In specific, we employ an U-Net architecture for discriminator, which is trained to predict the generator feature maps given fake images as inputs. Extensive experiments on mulitple datasets show that our GGDR consistently improves the performance of baseline methods in terms of quantitative and qualitative aspects. Code is available at https://github.com/naver-ai/GGDR