1 paper
Ci-Hao Wu, Tian-Sheuan Chang
Transformer-based speech enhancement models yield impressive results. However, their heterogeneous and complex structure restricts model compression potential, resulting in greater…