Humans are emotional beings, and emotions are one of the main drivers of human thoughts and actions. Therefore, for all environments designed for humans, it is essential that emotion processing capabilities such as analysis, recognition, and synthesis be incorporated. Naturally, any type of information, such as audio, visual, written, mental, or physiological, can be used for these tasks.

In this chapter, our concentration will be on emotion recognition from speech. Specifically, this chapter discusses the collection and organization of databases and emotional descriptors; the calculation, selection, and normalization of relevant speech features; and the models used to recognize emotions. We outline achievements, open questions, and future challenges in building Effective Automatic Speech Emotion Recognition (EASER) systems. It is known that emotions cause mental and physiological changes that also reflect in uttered speech.