Time-Frequency Masking in the Complex Domain for Speech Dereverberation and Denoising
Citations
218 citations
Cites background from "Time-Frequency Masking in the Compl..."
...Both two phase-based masks were demonstrated to be more effective than normal IRMs in suppressing the reverberated noise in Williamson and Wang (2017b)....
[...]
...For this reason, Williamson and Wang (2017b) further developed this approach, naming it complex IRM (cIRM)....
[...]
206 citations
152 citations
150 citations
89 citations
Cites background from "Time-Frequency Masking in the Compl..."
...hat it outperforms the phase-nonsensitive approaches. Note that, PSM does not completely enhance reverberant speech, since it cannot completely restore the phase. For this reason, Williamson and Wang [95] further developed this approach, naming it complex IRM (cIRM). It is defined as Mc(n,f)= |S(n,f)| |Y (n,f)| ej(θs−θy). (19) Therefore, cIRM can be regarded as the IRM in the complex domain, while PSM ...
[...]
References
7,244 citations
6,984 citations
"Time-Frequency Masking in the Compl..." refers methods in this paper
...Adaptive gradient descent [7] with a momentum term is used....
[...]
[...]
5,181 citations
3,720 citations
"Time-Frequency Masking in the Compl..." refers methods in this paper
...Simulated RIRs are generated using the imaging method [1], which is implemented in [12]....
[...]
2,969 citations
"Time-Frequency Masking in the Compl..." refers background in this paper
...These features include amplitude modulation spectrogram (AMS) [23], relative spectral transform and perceptual linear prediction (RASTA-PLP) [14], [15], mel-frequency cepstral coefficients (MFCC), as well as their deltas....
[...]
...[22] B. E. D. Kingsbury and N. Morgan, “Recognizing reverberant speech with RASTA-PLP,” in Proc....
[...]