Back there with another question, have anyone considered disentangling/learning on style of temporal structure in videos ?

I have seen many examples of ‘artistic’/photorealistic style transfer usage on single images/frames, and also using a temporal constraint to transfer it coherently.. But i ain’t askin’ about spatial style transfer enforced by temporal structure, but rather exploiting the temporality of different videos (films would be very useful i can imagine) and transferring ’em..