Acoustic Model with Multiple Lexicon Types for Indonesian Speech Recognition

Table 7

Three different sizes of training utterances.

Dataset	Number of utterances
Training	10,000	30,000	46,550
Testing	2,450	2,450	2,450