Cost-Sensitive Learning for Emotion Robust Speaker Recognition

Figure 4

DET curves for the traditional speaker models trained with neutral speech only.