Spoken Language Identification Using Deep Learning

Table 5

Dataset description.

Dataset	Spoken language identification [30]	Language identification dataset [31]	Common voice Kaggle dataset [32]	Mozilla common voice dataset [33]
Number of languages	3	22	16	4
Total samples	Train = 73080 (420 mins) Test = 540 (90 mins)	22000	354785	23842
Type	Audio	Text	Audio	Audio and TSV
Length	10 seconds	7 to 10 sentences in each line	Less than 10 seconds	Less than 10 seconds
Extension	FLAC	CSV	Mp3	Mp3 and TSV