S. Yatawatta

Recovering Weak Signals with Normalizing Flows

This work proposes the use of normalizing flow models to reconstruct calibration-affected weak signals by leveraging the statistical invariance of the target signals and assuming minimal initial suppression, and effectively recovers the lost signal components.