A High-Fidelity Speech Super Resolution Network using a Complex Global Attention Module with Spectro-Temporal Loss

Open in new window