Arbitrary Style Transfer using Graph Instance Normalization
This work addresses style transfer for image synthesis, offering a novel method that enhances robustness and applicability to related tasks like image-to-image translation, though it is incremental over existing normalization techniques.
The paper tackles the problem of arbitrary style transfer by introducing Graph Instance Normalization (GrIN), a learnable normalization technique using graph convolutional networks, which improves robustness by considering relationships between features, achieving competitive results on style transfer tasks.
Style transfer is the image synthesis task, which applies a style of one image to another while preserving the content. In statistical methods, the adaptive instance normalization (AdaIN) whitens the source images and applies the style of target images through normalizing the mean and variance of features. However, computing feature statistics for each instance would neglect the inherent relationship between features, so it is hard to learn global styles while fitting to the individual training dataset. In this paper, we present a novel learnable normalization technique for style transfer using graph convolutional networks, termed Graph Instance Normalization (GrIN). This algorithm makes the style transfer approach more robust by taking into account similar information shared between instances. Besides, this simple module is also applicable to other tasks like image-to-image translation or domain adaptation.