Audio-visual speech recognition on MuAViC Clean environment (test) benchmark leaderboard · SOTA2 Research

Audio-visual speech recognition on MuAViC Clean environment (test)

2.5En AccXLS-R 300M

Evaluation Results

Method	Links
XLS-R 300M2024.03		2.5	85.6	44	34.4	13.2	15.1	14.3	16.2	34.4	28.8
XLAVS-R 300M2024.03		2.5	82.4	45.6	24.2	11.3	14.6	12.9	13.7	33	26.7
XLAVS-R 300M2024.03		2.4	81.7	44.7	24.3	10.9	14.4	12.8	13.2	32.7	26.3
AV-HuBERT2024.03		2	89.6	53.5	47.2	18.2	20.7	21.2	22.9	46	35.7
u-HuBERT2024.03		1.9	89.3	53.3	47.2	17.9	20.7	21.6	22.5	45.8	35.6
u-HuBERT2024.03		1.9	89.3	52.1	46.4	17.3	20.5	21.2	21.9	44.4	35
XLAVS-R 2B2024.03		1.9	80.3	45.5	19.5	9.4	12.5	10.9	11.6	26	24.2
AV-HuBERT2024.03		1.7	89.4	52	46.2	17.4	20.3	20.8	22.1	44.7	35
XLAVS-R 2B2024.03		1.7	79.3	44.4	19	9.1	12.3	10.6	11.2	25	23.6