Multimodal Sparse Reconstruction Via Learning Cross-Modality MapsMS23

We introduce a framework to reconstruct undersampled signals with the aid of correlated signals. The correlated signals can arise from a different modality and, thus, need not be similar to the target signals and can have different representations. Our framework has two elements: L1-L1 minimization theory, which allows reconstructing undersampled signals with the aid of similar (same modality) signals, and an approach to learn mappings between modalities. Guarantees are obtained via statistical learning theory.

This presentation is part of Minisymposium “MS23 - Multi-Modality/Multi-Spectral Imaging and Structural Priors (2 parts)
organized by: Matthias J. Ehrhardt (University of Cambridge) , Simon Arridge (University College London) .

Joao Mota (Heriot-Watt University)