Self-supervision versus synthetic datasets: which is the lesser evil in the context of video denoising? | Publicación