Figure 2From: On the Impact of Children's Emotional Speech on Acoustic and Language ModelsDistribution of the prototypicality of the chunks in the four training sets Ohm_N, Ohm_M, Ohm_E, and Ohm_A. The average level of prototypicality is 0.79 for Ohm_N, 0.61 for Ohm_M, and 0.62 for both Ohm_E and Ohm_A.Back to article page