No title
In recent years, machine learning models have changed the field of video denoising. Most state-of-the-art models utilize multiple frames as input to denoise a single frame, which limits the temporal context window and often requires more computational power. Another, less explored, approach is to use recurrence to gain a hidden memory state in which temporal information can be stored. While the fo
