Speaker Characteristics and Emotion Classification

In this paper, we address the — interrelated — problems of speaker characteristics (personalization) and suboptimal performance of emotion classification in state-of-the-art modules from two different points of view: first, we focus on a specific phenomenon (irregular phonation or laryngealization) and argue that its inherent multi-functionality and speaker-dependency makes its use as feature in emotion classification less promising than one might expect. Second, we focus on a specific application of emotion recognition in a voice portal and argue that constraints on time and budget often prevent the implementation of an optimal emotion recognition module.