Dense CNN with Self-Attention for Time-Domain Speech Enhancement

Ashutosh Pandey and DeLiang Wang

This webpage provides sound demos for the algorithm described in the following paper:

Pandey A. and Wang D.L. (2021): Dense CNN with self-attention for time-domain speech enhancement. IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 29, pp. 1270-1279.


Babble noise, SNR = -5 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3



Babble noise, SNR = 0 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3



Babble noise, SNR = 5 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3



Cafeteria noise, SNR = -5 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3



Cafeteria noise, SNR = 0 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3



Cafeteria noise, SNR = 5 dB
Id Noisy DCN-T DCN-SM DCN-TF DCN-PCM Clean
F1
F2
F3
M1
M2
M3