Generating Artistic Styles using Neural Style Transfer
Esha Ghorpade, Nitisha Pradhan, Rahul Pal · Zenodo (CERN European Organization for Nuclear Research) · 2021
Neural style transfer is an optimization technique used to take two images—a content image and a style reference image (such as an artwork by a famous painter)—and blend them together so the output image looks like the content image, but gives the result such that it seems painted in the style of the style reference image. Style transfer that we intend on showcasing in this paper is an extremely fun and creative concept which puts forth the capabilities and internal representations of neural networks particularly convolutional neural networks(CNN). Style Transfer is a classic example of image stylization which is basically an image processing and manipulation technique. This paper makes the use of a Convolutional Neural Network called VGG16 to achieve this task. Transfer learning approach is used in our paper wherein VGG16 is used as the network from which the content and style outputs are obtained which give the output by capturing style of the style image and transferring it over the content image.