High-Quality Self-Supervised Deep Image Denoising
This enables denoising in scenarios where acquiring clean or paired data is expensive or impossible, addressing a practical bottleneck in image processing.
The paper tackles the problem of training image denoising models without clean reference images, achieving results comparable to state-of-the-art neural network denoisers for i.i.d. additive Gaussian noise and competitive performance for Poisson and impulse noise.
We describe a novel method for training high-quality image denoising models based on unorganized collections of corrupted images. The training does not need access to clean reference images, or explicit pairs of corrupted images, and can thus be applied in situations where such data is unacceptably expensive or impossible to acquire. We build on a recent technique that removes the need for reference data by employing networks with a "blind spot" in the receptive field, and significantly improve two key aspects: image quality and training efficiency. Our result quality is on par with state-of-the-art neural network denoisers in the case of i.i.d. additive Gaussian noise, and not far behind with Poisson and impulse noise. We also successfully handle cases where parameters of the noise model are variable and/or unknown in both training and evaluation data.