Progressive Growing of GANs for Improved Quality | PGGAN (paper illustrated)

Progressive Growing of GANs for Improved Quality | PGGAN

Progressive GAN was an ICLR 2018 oral paper. The idea of simply growing both the generator and discriminator networks of GANs proves to be quite effective at generating very high resolution images at 1024X1024. In addition to generating mind blowing images, the paper also proposes some additional nuances to training GANs like Pixel feature vector normalization, Equalized learning rate and Minibatch Standard Deviation. This video is all about Progressive GAN - its architecture, the ideas and results.

Abstract:
We describe a new training methodology for generative adversarial networks. The key idea is to grow both the generator and discriminator progressively: starting from a low resolution, we add new layers that model increasingly fine details as training progresses. This both speeds the training up and greatly stabilizes it, allowing us to produce images of unprecedented quality, e.g., CelebA images at 1024². We also propose a simple way to increase the variation in generated images, and achieve a record inception score of 8.80 in unsupervised CIFAR10. Additionally, we describe several implementation details that are important for discouraging unhealthy competition between the generator and discriminator. Finally, we suggest a new metric for evaluating GAN results, both in terms of image quality and variation. As an additional contribution, we construct a higher-quality version of the CelebA dataset.

#deeplearning #machinelearning #aibites

[AICoffeeBreak]

Wow, the audio quality increased dramatically! Keep going! 💪