Controllable Feature-Preserving Style Transfer

This paper proposes a new style transfer quality assessment approach introducing quantifiable metrics to optimize. First, we utilize a pre-trained DualStyleGAN model to generate multiple stylized portraits in the style vector space. Then, we design a custom scoring mechanism that uses the newly proposed CSCI and CCVI metrics to evaluate the results’ structural similarity, color consistency, and edge retention. We select and optimize the top outputs using human aesthetic standards to obtain the most natural, beautiful, and artistic results. Experimental results show that our proposed evaluation pipeline can effectively improve the quality of style transfer.