Self-supervised learning for phase retrieval
This addresses a bottleneck in medical and scientific imaging by enabling phase retrieval without fully sampled data, though it appears incremental as it builds on prior self-supervised advances.
The paper tackles the problem of phase retrieval in imaging, where supervised learning is limited by the lack of fully sampled data, and proposes a self-supervised method that leverages image translation invariance to achieve reconstruction without ground truth references.
In recent years, deep neural networks have emerged as a solution for inverse imaging problems. These networks are generally trained using pairs of images: one degraded and the other of high quality, the latter being called 'ground truth'. However, in medical and scientific imaging, the lack of fully sampled data limits supervised learning. Recent advances have made it possible to reconstruct images from measurement data alone, eliminating the need for references. However, these methods remain limited to linear problems, excluding non-linear problems such as phase retrieval. We propose a self-supervised method that overcomes this limitation in the case of phase retrieval by using the natural invariance of images to translations.