arXiv 2111.15624
Image Style Transfer and Content-Style Disentanglement
By Sailun Xu, Jiazhi Zhang, et al.
Published 2021-11-25
Mindmap
Browse the paper's core ideas, clusters, and relationships in a structured outline.
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 reco…