Acoustic Model with Multiple Lexicon Types for Indonesian Speech Recognition

Table 10

The performance of the acoustic models using 30,000 validated utterances.

Lexicon type	GMM-HMM model (%WER)
	Monophone	Triphone DELTA + DELTA-DELTA	Triphone LDA + MLLT	Triphone SAT
1c-vocab-full	81.75	65.41	54.13	54.90
enmap-1c-vocab-full	81.34	65.23	54.00	55.09
enmap-kv-vocab-full	80.76	71.21	61.66	62.55
kv-vocab-full	77.16	69.85	60.76	61.73