CapsuleGAN: Generative Adversarial Capsule Network
This work addresses image generation and classification for machine learning researchers, presenting an incremental improvement by integrating capsule networks into GANs.
The paper tackles the problem of modeling image data distribution by replacing standard convolutional neural networks with capsule networks as discriminators in generative adversarial networks, resulting in CapsuleGAN outperforming convolutional-GAN on MNIST and CIFAR-10 datasets in generative adversarial metrics and semi-supervised classification.
We present Generative Adversarial Capsule Network (CapsuleGAN), a framework that uses capsule networks (CapsNets) instead of the standard convolutional neural networks (CNNs) as discriminators within the generative adversarial network (GAN) setting, while modeling image data. We provide guidelines for designing CapsNet discriminators and the updated GAN objective function, which incorporates the CapsNet margin loss, for training CapsuleGAN models. We show that CapsuleGAN outperforms convolutional-GAN at modeling image data distribution on MNIST and CIFAR-10 datasets, evaluated on the generative adversarial metric and at semi-supervised image classification.