List your company

Scene-aware visually driven speech synthesis on Vivid-210K (test)

7.15WERVividVoice
7.01127.94818.8859.8219Feb 1, 2026

Evaluation results

MethodLinks
VividVoice2026.027.153.981.530.253.953.084.33.88
VoiceLDM2026.029.234.741.790.273.231.752.563.41
GT2026.0210.62——0.394.364.034.114.25