VISinger-2

VISinger2 is a benchmark dataset used to evaluate Singing Voice Synthesis (SVS) systems, focusing on enhancing the quality of synthesized singing voices by leveraging both labeled and unlabeled data.

Papers using VISinger-2 (3)

Sifisinger: A High-fidelity End-to-end Singing Voice Synthesizer Based On Source-filter Model2024 · 3 citesVisinger2+: End-to-end Singing Voice Synthesis Augmented By Self-supervised Learning Representation2024 · 3 citesVISinger2+: End-to-End Singing Voice Synthesis Augmented by Self-Supervised Learning Representation2024