Single Channel Speech Enhancement Using Adaptive Soft-Thresholding with Bivariate EMD

Figure 5

Segmental SNR using bEMD and uEMD-based denoising techniques for noisy speech corrupted by white noise—0 dB SNR (a) and 5 dB SNR (b).

(a)

(b)