Violin Teaching Improvement Strategy in the Context of Intelligent Internet of Things

Table 2

PESQ evaluation of speech separation models under different training times.

SNR (dB)	PESQ_voice_40	PESQ_voice_50	PESQ_voice_60
−5	2.58	2.73	2.60
0	2.94	2.91	2.88
5	3.13	3.25	3.23
10	3.26	3.46	3.35
15	3.45	3.58	3.57
20	3.62	3.75	3.73
Avg.	3.18	3.32	3.27