Spoken Term Detection on LibriSpeech OOV benchmark leaderboard · SOTA2 Research

Spoken Term Detection on LibriSpeech OOV

0.02MTWV (-5dB)HuBERT-Large

Evaluation Results

Method	Links
HuBERT-Large2025.12		0.02	0.07	0.12	0.2	0.26	0.29
SpeechTokenizer2025.12		0.02	0.04	0.06	0.11	0.13	0.16
WavLM-Large2025.12		0.04	0.07	0.14	0.21	0.27	0.31
SpeechTokenizer2025.12		0.13	0.21	0.3	0.42	0.48	0.49
BEST-STD2025.12		0.13	0.2	0.27	0.33	0.39	0.43
WavLM-Large2025.12		0.15	0.22	0.29	0.31	0.35	0.37
HuBERT-Large2025.12		0.16	0.27	0.34	0.4	0.41	0.43
WavLM-Large2025.12		0.17	0.25	0.35	0.43	0.47	0.49
BEST-STD2025.12		0.22	0.29	0.37	0.44	0.49	0.54
WavLM-Large2025.12		0.29	0.37	0.41	0.42	0.43	0.45
Ours-Transformer2025.12		0.4	0.46	0.52	0.55	0.57	0.57
BEST-STD 2.02025.12		0.4	0.5	0.56	0.58	0.61	0.62
Ours-Transformer2025.12		0.5	0.56	0.6	0.62	0.64	0.65
BEST-STD 2.02025.12		0.51	0.62	0.65	0.67	0.68	0.68