Creating various styles of animations using example-based filtering

A number of nonphotorealistic rendering methods for producing artistic style images have been developed. Recently, a method called "Image Analogies" was proposed. This method is based on the idea of example-based filtering that uses a pair of images (an original image and a filtered image) as training data and creates an image which has a style that resembles the filtered image of the training data. This method has great flexibility since the users can create various styles of images by simply changing the training data. We extend "Image Analogies" to create various styles of animations. The coherence between frames is considered by computing the pixel flows between the frames. From experimental results, the proposed method can create nice animation sequences.