Artistic style transfer for videos

Manuel Ruder, Alexey Dosovitskiy, Thomas Brox

German Conference on Pattern Recognition (GCPR), LNCS 9796, pp. 26-36 (2016)

Paper: https://arxiv.org/abs/1604.08610
Repository1 (original): https://github.com/manuelruder/artistic-videos
Repository2 (faster): https://github.com/NameRX/artistic-videos

Deep learning based style transfer, extended to videos. Frame-to-frame coherence is guided by optical flow.
Input is a video (content) and an image (style) to output stylized image.

カテゴリー: Blog