Figure 9From: A sub-band-based feature reconstruction approach for robust speaker recognition SNR-dependent speaker recognition performance in babble and factory1 noise. Back to article page