Lensless computational imaging through deep learning
This work addresses the challenge of imaging without lenses, which could benefit fields like microscopy and portable devices, though it appears incremental as it applies existing deep learning methods to a new imaging context.
The paper tackled the problem of lensless computational imaging by using deep neural networks to solve inverse problems, demonstrating for the first time that DNNs can recover phase objects from raw intensity images recorded at a distance.
Deep learning has been proven to yield reliably generalizable answers to numerous classification and decision tasks. Here, we demonstrate for the first time, to our knowledge, that deep neural networks (DNNs) can be trained to solve inverse problems in computational imaging. We experimentally demonstrate a lens-less imaging system where a DNN was trained to recover a phase object given a raw intensity image recorded some distance away.