Model-blind Video Denoising Via Frame-to-frame Training
This addresses the challenge of denoising videos from unknown sources for applications in video processing, though it is incremental as it builds on existing denoising networks.
The authors tackled the problem of video denoising without prior knowledge of the camera's processing chain by proposing a blind method that fine-tunes a pre-trained network using a novel frame-to-frame training strategy, achieving state-of-the-art performance for Gaussian noise and visually-pleasing results for various perturbations.
Modeling the processing chain that has produced a video is a difficult reverse engineering task, even when the camera is available. This makes model based video processing a still more complex task. In this paper we propose a fully blind video denoising method, with two versions off-line and on-line. This is achieved by fine-tuning a pre-trained AWGN denoising network to the video with a novel frame-to-frame training strategy. Our denoiser can be used without knowledge of the origin of the video or burst and the post processing steps applied from the camera sensor. The on-line process only requires a couple of frames before achieving visually-pleasing results for a wide range of perturbations. It nonetheless reaches state of the art performance for standard Gaussian noise, and can be used off-line with still better performance.