Deep Learning-Based Amplitude Fusion for Speech Dereverberation

Table 10

of various methods under different levels of reverberation.

MAP	DCC	MAP + DCC
Far field	S	4.06	3.00	2.57
	M	3.75	3.74	2.86
	L	3.86	4.05	2.89
Near field	S	4.15	2.94	2.61
	M	3.87	3.18	2.66
	L	3.66	4.40	3.06
AVE		3.89	3.55	2.78