Deep Image Style Transfer from Freeform Text

TL;DR

This paper introduces a novel deep neural style transfer method that generates style images from freeform text descriptions, enabling more flexible and high-quality image stylization.

Contribution

It presents a seamless pipeline combining language and style transfer models to create style images from text, improving style transfer quality over baseline methods.

Findings

Generated style images closely match text descriptions

Output images have similar losses and better quality

The method demonstrates effective integration of language and style transfer models

Abstract

This paper creates a novel method of deep neural style transfer by generating style images from freeform user text input. The language model and style transfer model form a seamless pipeline that can create output images with similar losses and improved quality when compared to baseline style transfer methods. The language model returns a closely matching image given a style text and description input, which is then passed to the style transfer model with an input content image to create a final output. A proof-of-concept tool is also developed to integrate the models and demonstrate the effectiveness of deep image style transfer from freeform text.