Figure 8From: Monaural Voiced Speech Segregation Based on Dynamic Harmonic FunctionSegmentation comparison. The input signal is a voiced speech mixed by click noise. (a) Segments formed by cross-channel correlation and time continuity. The black region is dominated by speech and the gray region is dominated by click noise. (b) Segments formed by cross-channel correlation, time continuity and carrier-to-envelope energy ratio.Back to article page