Speech-to-Portrait on VoxCeleb (test) benchmark leaderboard · SOTA2 Research

Speech-to-Portrait on VoxCeleb (test)

25.24L1 ErrorSCLDM

Evaluation Results

Method	Links
SCLDM2025.10		25.24	0.91	9.86	99.6	89.3	19.84	33.57	51.79
SCLDM2025.10		25.24	0.91	0.0986	99.6	89.3	19.84	33.57	51.79
Kato et al.2025.10		40.11	2.26	18.74	98.1	83.8	16.19	25.64	42.38
Kato et al.2025.10		40.11	2.26	0.1874	98.1	83.8	16.19	25.64	42.38
Choi et al.2025.10		56.32	2.24	30.49	97.6	74.8	9.43	16.32	28.67
Choi et al.2025.10		56.32	2.24	0.3049	97.6	74.8	9.43	16.32	28.67
Wen et al.2025.10		59.82	2.41	42.54	97.4	72.5	8.26	15.62	23.51
Wen et al.2025.10		59.82	2.41	0.4254	97.4	72.5	8.26	15.62	23.51
Speech2Face2025.10		66.46	2.77	44.38	96.1	69.4	7.79	14.38	20.14
Speech2Face2025.10		66.46	2.77	0.4438	96.1	69.4	7.79	14.38	20.14
SF2F2025.10		78.45	13.31	58.79	79.3	57.6	9.25	17.17	22.53
SF2F2025.10		78.45	13.31	0.5879	79.3	57.6	9.25	17.17	22.53
Wav2Pix2025.10		137.58	22.19	79.36	74.5	49.6	4.81	9.56	12.94
Wav2Pix2025.10		137.58	22.19	0.7936	74.5	49.6	4.81	9.56	12.94