The learning to synthesize method produces high-resolution, artifact-free phase reconstructions resilient to low photon flux by separately learning low and high frequency bands and then synthesizing them.
Learning to synthesize: splitting and recombining low and high spatial frequencies for image recovery
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
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.
fields
eess.IV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning to Synthesize: Robust Phase Retrieval at Low Photon counts
The learning to synthesize method produces high-resolution, artifact-free phase reconstructions resilient to low photon flux by separately learning low and high frequency bands and then synthesizing them.