Style Transfer Dataset: What Makes A Good Stylization?
This work addresses the need for better evaluation methods in style transfer research, providing a dataset and insights that can automate tasks like configuration and evaluation, though it is incremental as it builds on existing style transfer work.
The authors tackled the problem of evaluating image style transfer by creating a dataset of 10,000 stylizations with manual ratings, identifying factors that influence user evaluations and showing quantitative measures with statistically significant impact.
We present a new dataset with the goal of advancing image style transfer - the task of rendering one image in the style of another image. The dataset covers various content and style images of different size and contains 10.000 stylizations manually rated by three annotators in 1-10 scale. Based on obtained ratings, we find which factors are mostly responsible for favourable and poor user evaluations and show quantitative measures having statistically significant impact on user grades. A methodology for creating style transfer datasets is discussed. Presented dataset can be used in automating multiple tasks, related to style transfer configuration and evaluation.