Hybrid dual-path network: Singing voice separation in the waveform domain by combining Conformer and Transformer architectures

Autor: Wang, Chunxi a, Jia, Maoshen a, ⁎, Li, Meiran a, Ma, Yong b, Yao, Dingding c
Zdroj: In Speech Communication March 2025 168
Databáze: ScienceDirect