Cued Speech-to-Speech Synthesis on Unified-HI Normal-hearing cuers v1 (test) benchmark leaderboard · SOTA2 Research

Cued Speech-to-Speech Synthesis on Unified-HI Normal-hearing cuers v1 (test)

4.135LSE-CCMML

Evaluation Results

Method	Links
CMML2025.06		4.135	0.663	9.241	1.24	11
Lip2Speech2025.06		4.215	0.803	9.367	1.03	5
EcoCued2025.06		4.327	0.657	9.146	1.28	12
LipVoicer2025.06		4.361	0.754	9.226	1.12	8
CSR2025.06		4.874	0.186	9.125	2.53	57
CSV2S2025.06		6.245	0.374	7.962	2.27	42
UniCUE2025.06		6.729	0.205	7.632	2.46	53
GT2025.06		7.274	—	7.314	2.79	—