Real-time coherent diffraction inversion using deep generative networks
This enables real-time imaging for applications like CDI, though it is incremental as it applies an existing deep learning method to a known bottleneck.
The paper tackles the problem of slow and computationally expensive iterative phase retrieval algorithms in coherent diffraction imaging (CDI) by introducing CDI NN, a deep generative network that predicts structure and phase from diffraction intensities. The result is real-time imaging with inversion times of a few milliseconds on a standard desktop machine.
Phase retrieval, or the process of recovering phase information in reciprocal space to reconstruct images from measured intensity alone, is the underlying basis to a variety of imaging applications including coherent diffraction imaging (CDI). Typical phase retrieval algorithms are iterative in nature, and hence, are time-consuming and computationally expensive, precluding real-time imaging. Furthermore, iterative phase retrieval algorithms struggle to converge to the correct solution especially in the presence of strong phase structures. In this work, we demonstrate the training and testing of CDI NN, a pair of deep deconvolutional networks trained to predict structure and phase in real space of a 2D object from its corresponding far-field diffraction intensities alone. Once trained, CDI NN can invert a diffraction pattern to an image within a few milliseconds of compute time on a standard desktop machine, opening the door to real-time imaging.