Visible to the public Incorporating Multiscale Contextual Loss for Image Style Transfer

TitleIncorporating Multiscale Contextual Loss for Image Style Transfer
Publication TypeConference Paper
Year of Publication2018
AuthorsLi, P., Zhao, L., Xu, D., Lu, D.
Conference Name2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC)
Date PublishedJune 2018
PublisherIEEE
ISBN Number978-1-5386-4991-6
Keywordscontent image, convolution, convolutional neural networks, distortion, feature extraction, feedforward neural nets, Haar loss, Haar transforms, high-level CNN features, image classification, image representation, Image resolution, image style transfer, image wavelet transform, lower-level CNN features, Metrics, multiscale contextual loss, neural style transfer, Optimization, optimization framework, Predictive Metrics, pubcrawl, resilience, Resiliency, Scalability, semantic information, Semantics, wavelet coefficients, wavelet transform, wavelet transforms
Abstract

In this paper, we propose to impose a multiscale contextual loss for image style transfer based on Convolutional Neural Networks (CNN). In the traditional optimization framework, a new stylized image is synthesized by constraining the high-level CNN features similar to a content image and the lower-level CNN features similar to a style image, which, however, appears to lost many details of the content image, presenting unpleasing and inconsistent distortions or artifacts. The proposed multiscale contextual loss, named Haar loss, is responsible for preserving the lost details by dint of matching the features derived from the content image and the synthesized image via wavelet transform. It endows the synthesized image with the characteristic to better retain the semantic information of the content image. More specifically, the unpleasant distortions can be effectively alleviated while the style can be well preserved. In the experiments, we show the visually more consistent and simultaneously well-stylized images generated by incorporating the multiscale contextual loss.

URLhttps://ieeexplore.ieee.org/document/8492852
DOI10.1109/ICIVC.2018.8492852
Citation Keyli_incorporating_2018