Audio-Driven Facial Animation on MEAD 41 (test) benchmark leaderboard · SOTA2 Research

Audio-Driven Facial Animation on MEAD 41 (test)

31.873PSNRESGaussianFace

Evaluation Results

Method	Links
ESGaussianFace2026.01		31.873	0.901	16.933	0.028	6.216	2.832	75.944
ESGaussianFace2026.01		31.873	0.901	16.933	0.028	6.216	2.832	75.944
Style2Talker2026.01		29.455	0.795	25.282	0.035	5.294	3.007	70.273
Style2Talker2026.01		29.455	0.795	25.282	0.035	5.294	3.007	70.273
GaussianTalker2026.01		28.911	0.818	22.29	0.038	5.202	4.981	52.621
GaussianTalker2026.01		28.911	0.818	22.29	0.038	5.202	4.981	52.621
SyncTalk2026.01		28.12	0.803	32.959	0.04	4.77	5.071	48.245
SyncTalk2026.01		28.12	0.803	32.959	0.04	4.77	5.071	48.245
MEAD2026.01		28.081	0.712	21.273	0.039	5.784	3.564	64.92
MEAD2026.01		28.081	0.712	21.273	0.039	5.784	3.564	64.92
EDTalk2026.01		27.938	0.76	34.497	0.048	5.436	4.467	62.288
EDTalk2026.01		27.938	0.76	34.497	0.048	5.436	4.467	62.288
Wav2Lip2026.01		27.819	0.67	46.373	0.051	5.851	5.068	16.802
Wav2Lip2026.01		27.819	0.67	46.373	0.051	5.851	5.068	16.802
DreamTalk2026.01		27.801	0.732	34.239	0.051	5.56	4.693	61.81
DreamTalk2026.01		27.801	0.732	34.239	0.051	5.56	4.693	61.81
EAT2026.01		27.65	0.69	35.094	0.049	5.277	4.36	58.485
EAT2026.01		27.65	0.69	35.094	0.049	5.277	4.36	58.485
MakeItTalk2026.01		27.597	0.602	53.09	0.059	4.495	5.208	19.379
MakeItTalk2026.01		27.597	0.602	53.09	0.059	4.495	5.208	19.379
EAMM2026.01		26.807	0.681	58.334	0.063	3.457	4.578	25.415
EAMM2026.01		26.807	0.681	58.334	0.063	3.457	4.578	25.415
Audio2Head2026.01		26.529	0.623	63.287	0.07	3.342	6.681	19.915
Audio2Head2026.01		26.529	0.623	63.287	0.07	3.342	6.681	19.915
AD-NeRF2026.01		25.282	0.549	84.506	0.086	1.298	—	10.245
AD-NeRF2026.01		25.282	0.549	84.506	0.086	1.298	—	10.245