ResearchBenchmarksListening Video Generation on VoxHear User Study 1.0 (test)Listening Video Generation on VoxHear User Study 1.0 (test)Follow4.14NaturalnessOurs1.30082.03792.7753.5121Apr 11, 2026Evaluation resultsMethodMethodLinksNaturalnessMotion DiversityAV AlignmentVisual QualityOurs2026.044.144.054.184.32INFP2026.043.8644.054.55DIM2026.041.682.0521.86L2L2026.041.591.451.771.73RLHG2026.041.411.361.681.5