Optimizing Shoulder to Shoulder: A Coordinated Sub-Band Fusion Model for Real-Time Full-Band Speech Enhancement

Audio Samples on DNS-Challenge 2022 (48kHz)

Sample Name	Noisy	NSNet2	SF-Net
A1CH3TODZNQCES_M_Breathing
A1CH3TODZNQCES_M_Clatter
A1CH3TODZNQCES_M_Cutlery
A1CH3TODZNQCES_M_Fan
A1CH3TODZNQCES_M_MouseClicks
A1CIM28ZUCA8RX_M_Breathing
A1CIM28ZUCA8RX_M_Baby
A1CH3TODZNQCES_M_Water
A1CIM28ZUCA8RX_M_Door

Audio Samples on VoiceBank + DEMAND dataset (48kHz)

Sample Name	Noisy	DS-Net(full)	CTS-Net(full)	SF-Net
p232_003
p232_005
p232_006
p232_007
p232_010
p232_011
p232_013
p232_015
p232_016
p232_017