Keywords: imaging, inverse problems, subspace projections, random Delaunay triangulations, CNN, geophysics, regularization
TL;DR: We solve ill-posed inverse problems with scarce ground truth examples by estimating an ensemble of random projections of the model instead of the model itself.
Abstract: We propose a new learning-based approach to solve ill-posed inverse problems in imaging. We address the case where ground truth training samples are rare and the problem is severely ill-posed---both because of the underlying physics and because we can only get few measurements. This setting is common in geophysical imaging and remote sensing. We show that in this case the common approach to directly learn the mapping from the measured data to the reconstruction becomes unstable. Instead, we propose to first learn an ensemble of simpler mappings from the data to projections of the unknown image into random piecewise-constant subspaces. We then combine the projections to form a final reconstruction by solving a deconvolution-like problem. We show experimentally that the proposed method is more robust to measurement noise and corruptions not seen during training than a directly learned inverse.
Code: [![github](/images/github_icon.svg) swing-research/deepmesh](https://github.com/swing-research/deepmesh)
Data: [CelebA](https://paperswithcode.com/dataset/celeba), [LSUN](https://paperswithcode.com/dataset/lsun)
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 1 code implementation](https://www.catalyzex.com/paper/arxiv:1805.11718/code)