Acoustic Model with Multiple Lexicon Types for Indonesian Speech Recognition

Table 3

The Indonesian audio dataset before validation.

Category	Original audio		Utterances
	Number of audio	Total duration (hours)	Number of utterances	Total duration (hours)
Autos and vehicles	20	2.8039	2,234	1.3289
Comedy	40	7.7125	6,017	4.0768
Education	553	54.9508	47,477	38.4886
Entertainment	308	57.1747	39,155	25.2972
Film and animation	77	7.7464	6, 475	4.5591
Gaming	1	0.1150	124	0.0987
Howto and style	355	54.5117	28,726	41.8189
Music	56	4.6469	1, 409	2.5783
News and politics	170	24.6003	19,295	16.8386
People and blogs	179	50.2578	32,205	22.7193
Pets and animals	2	0.3153	57	0.0433
Science and technology	215	44.5308	31,492	22.9787
Sports	1	0.1583	137	0.1131
Travel and events	1	0.2236	78	0.0633
Uncategorized	2	0.3367	410	0.2515
Total	1,980	310.0847	215,291	181.2543