FCC-GAN: A Fully Connected and Convolutional Net Architecture for GANs
This work addresses a fundamental architectural choice for researchers and practitioners in generative modeling, though it appears incremental as it modifies existing convolutional approaches.
The paper tackles the problem of unclear optimal network architectures for GANs in image synthesis by proposing FCC-GAN, which incorporates fully connected and pooling layers alongside convolutions, resulting in faster learning and higher-quality samples across four datasets.
Generative Adversarial Networks (GANs) are a powerful class of generative models. Despite their successes, the most appropriate choice of a GAN network architecture is still not well understood. GAN models for image synthesis have adopted a deep convolutional network architecture, which eliminates or minimizes the use of fully connected and pooling layers in favor of convolution layers in the generator and discriminator of GANs. In this paper, we demonstrate that a convolution network architecture utilizing deep fully connected layers and pooling layers can be more effective than the traditional convolution-only architecture, and we propose FCC-GAN, a fully connected and convolutional GAN architecture. Models based on our FCC-GAN architecture learn both faster than the conventional architecture and also generate higher quality of samples. We demonstrate the effectiveness and stability of our approach across four popular image datasets.