Japanese Animation Style Transfer Using Deep Neural Networks
Shiyang Ye, Ryo Ohtera · 2017 International Conference on Information, Communication and Engineering (ICICE) · 2017
This paper tries to find a deep-learning approach to Japanese Animation style transfer that creates animation background images more easily and efficiently so that we can reduce the cost and workload for Japanese animation industry. Our approach now builds is based on a deep-learning approach using Convolutional Neural Networks(CNN) which is most popular in image processing tasks. Moreover, we limit the number of the color used in the output images to get close to an ideal color.