A Study of Batik Style Transfer using Neural Network
Aditya Firman Ihsan · 2021
In this study, two remarkable applications of convolutional neural network, i.e. texture synthesis and style transfer are applied to batik texture. Individual layers from four pre-trained networks such as VGG-19, Inception V3, ResNet-50, and DenseNet-121 are compared and analyzed. Different batik motifs with some specific criteria are also compared to see the capability of original style transfer algorithm to regenerate concrete texture of Batik. Lastly, we propose a way to reconstruct batik images with some patterns following an object’s shape contained in content image.