Learning to synthesize: splitting and recombining low and high spatial frequencies for image recovery
This addresses image recovery issues in fields like microscopy and imaging, offering a method to enhance fidelity across frequency bands, though it is incremental as it builds on prior work.
The paper tackles the problem of uneven fidelity in DNN-based image reconstruction by proposing LS-DNN, which splits and recombines low and high spatial frequencies using separate DNNs, resulting in improved performance in super-resolution and quantitative phase retrieval tasks.
Deep Neural Network (DNN)-based image reconstruction, despite many successes, often exhibits uneven fidelity between high and low spatial frequency bands. In this paper we propose the Learning Synthesis by DNN (LS-DNN) approach where two DNNs process the low and high spatial frequencies, respectively, and, improving over [30], the two DNNs are trained separately and a third DNN combines them into an image with high fidelity at all bands. We demonstrate LS-DNN in two canonical inverse problems: super-resolution (SR) in diffraction-limited imaging (DLI), and quantitative phase retrieval (QPR). Our results also show comparable or improved performance over perceptual-loss based SR [21], and can be generalized to a wider range of image recovery problems.