Whitening and Coloring batch transform for GANs

Abstract

Batch Normalization (BN) is a common technique used to speed-up and stabilize training. On the other hand, the learnable parameters of BN are commonly used in conditional Generative Adversarial Networks (cGANs) for representing class-specific information using conditional Batch Normalization (cBN). In this paper we propose to generalize both BN and cBN using a Whitening and Coloring based batch normalization. We show that our conditional Coloring can represent categorical conditioning information which largely helps the cGAN qualitative results. Moreover, we show that full-feature whitening is important in a general GAN scenario in which the training process is known to be highly unstable. We test our approach on different datasets and using different GAN networks and training protocols, showing a consistent improvement in all the tested frameworks. Our CIFAR-10 conditioned results are higher than all previous works on this dataset.

arXiv commentICLR 2019

Cited by in corpus (9)

• Concept Whitening for Interpretable Image Recognition
• ContraGAN: Contrastive Learning for Conditional Image Generation
• Robust Differentiable SVD
• Whitening for Self-Supervised Representation Learning
• Multi-Channel Attention Selection GAN with Cascaded Semantic Guidance for Cross-View Image Translation
• Diversified Arbitrary Style Transfer via Deep Feature Perturbation
• On Feature Decorrelation in Self-Supervised Learning
• Group Whitening: Balancing Learning Efficiency and Representational Capacity
• An Investigation into the Stochasticity of Batch Whitening