Learning to Recorrupt: Noise Distribution Agnostic Self-Supervised Image Denoising
This addresses the challenge of denoising images under unknown noise conditions, which is incremental by eliminating the need for noise distribution knowledge in recorruption-based methods.
The paper tackles the problem of self-supervised image denoising without requiring prior knowledge of the noise distribution, achieving state-of-the-art performance across unconventional and heavy-tailed noise distributions like log-gamma, Laplace, spatially correlated, and Poisson-Gaussian noise.
Self-supervised image denoising methods have traditionally relied on either architectural constraints or specialized loss functions that require prior knowledge of the noise distribution to avoid the trivial identity mapping. Among these, approaches such as Noisier2Noise or Recorrupted2Recorrupted, create training pairs by adding synthetic noise to the noisy images. While effective, these recorruption-based approaches require precise knowledge of the noise distribution, which is often not available. We present Learning to Recorrupt (L2R), a noise distribution-agnostic denoising technique that eliminates the need for knowledge of the noise distribution. Our method introduces a learnable monotonic neural network that learns the recorruption process through a min-max saddle-point objective. The proposed method achieves state-of-the-art performance across unconventional and heavy-tailed noise distributions, such as log-gamma, Laplace, and spatially correlated noise, as well as signal-dependent noise models such as Poisson-Gaussian noise.