Application of Transfer Learning Algorithm and Real Time Speech Detection in Music Education Platform

Table 1

Speech recognition accuracy under different training hours.

Research objects	Training hours
	25	50	100	200
Female sex	29.30	36.19	40.60	41.30
Male	30.42	35.89	41.32	42.18
Female/male	22.24	29.32	29.80	30.87
Female/male (double)	37.82	43.21	45.26	46.32