dAUTOMAP: decomposing AUTOMAP to achieve scalability and enhance performance

TL;DR

dAUTOMAP is a scalable decomposition of AUTOMAP that improves performance and reduces parameters, making it more practical for generalized reconstruction tasks.

Contribution

The paper introduces dAUTOMAP, a novel decomposition method that enables AUTOMAP to scale linearly and outperform the original with fewer parameters.

Findings

dAUTOMAP outperforms AUTOMAP in reconstruction quality.

dAUTOMAP requires significantly fewer parameters.

The approach enhances scalability for generalized reconstruction.

Abstract

AUTOMAP is a promising generalized reconstruction approach, however, it is not scalable and hence the practicality is limited. We present dAUTOMAP, a novel way for decomposing the domain transformation of AUTOMAP, making the model scale linearly. We show dAUTOMAP outperforms AUTOMAP with significantly fewer parameters.