Analysis of Two-Piano Teaching Assistant Training Based on Neural Network Model Sound Sequence Recognition

Table 3

High signal-to-noise ratio model identification results.

Input noise type	Whitt (%)	F16 (%)	HF channel (%)
Noisy	18.6	14.8	17.8
AET	16.2	12.6	15.3
R-CED	13.5	10.5	12.8
UNetSMB	12.7	9.3	12.1