Figure 5From: Contribution of statistical tests to sparseness-based blind source separation Speech mixture spectrogram when mixing matrix A is applied to the four sources of Figure 4 (SNR = 10 dB). Back to article page