ResearchBenchmarksAudio-driven head video generation on PortraitOneMin headAudio-driven head video generation on PortraitOneMin headFollow8.602SCVASA-16.89127.335357.77958.22365Jun 1, 2026Evaluation resultsMethodMethodLinksSCSDCAPPFVD25FPSVASA-1Resolution=512 x 512,...Resolution=512 x 512, Hardware=RTX 4090 GPU2026.068.6026.5410.679119.47437.9Ours M=1Resolution=512 x 512,...Resolution=512 x 512, M (Number of reference images)=1, Hardware=RTX 4090 GPU2026.068.1826.820.601102.46823.3SadTalkerResolution=512 x 512,...Resolution=512 x 512, Hardware=RTX 4090 GPU2026.066.9577.7050.413270.6673.8