Image Style Transfer and Content-Style Disentanglement
This addresses the problem of flexible image style manipulation for computer vision applications, though it appears incremental as it builds on existing disentanglement and style transfer techniques.
The paper tackles the problem of disentangling content and style representations in images, enabling style extrapolation and interpolation between styles. The result is a method that ensures separation of content and style information through data augmentation and triplet loss, with cycle-consistency loss ensuring faithful image reconstruction.
We propose a way of learning disentangled content-style representation of image, allowing us to extrapolate images to any style as well as interpolate between any pair of styles. By augmenting data set in a supervised setting and imposing triplet loss, we ensure the separation of information encoded by content and style representation. We also make use of cycle-consistency loss to guarantee that images could be reconstructed faithfully by their representation.