UNSURE: self-supervised learning with Unknown Noise level and Stein's Unbiased Risk Estimate
This addresses a practical limitation in self-supervised learning for imaging, where noise levels are often unknown, making it incremental by improving robustness over prior methods.
The paper tackles the problem of self-supervised learning for image reconstruction without requiring known noise levels, proposing a new method based on Stein's Unbiased Risk Estimate that outperforms existing approaches on various imaging inverse problems.
Recently, many self-supervised learning methods for image reconstruction have been proposed that can learn from noisy data alone, bypassing the need for ground-truth references. Most existing methods cluster around two classes: i) Stein's Unbiased Risk Estimate (SURE) and similar approaches that assume full knowledge of the noise distribution, and ii) Noise2Self and similar cross-validation methods that require very mild knowledge about the noise distribution. The first class of methods tends to be impractical, as the noise level is often unknown in real-world applications, and the second class is often suboptimal compared to supervised learning. In this paper, we provide a theoretical framework that characterizes this expressivity-robustness trade-off and propose a new approach based on SURE, but unlike the standard SURE, does not require knowledge about the noise level. Throughout a series of experiments, we show that the proposed estimator outperforms other existing self-supervised methods on various imaging inverse problems.