Image Textural Style Transfer using Neural Network

Abstract

Abstract-In this survey paper, the purpose was to present the context aware algorithm based textural style transfer based on sporadic representation of texture fusion. The ideology behind this is to blend the style which is identified from a reference image to a source image, while the scenic attributes of the source image are conserved. The sporadic representation focuses on the extraction of the texture or style from a reference or example image(s). The actual fusion takes place of the extracted style component with the source image where the settings are optimized to conserve the true scene structure. Here, the textural style transfer is performed automatically without the required knowledge of user interest.