The aim of this book is to bring out various features through speech processing, and use them in an acoustic model to recognize the emotion conveyed by the person. The topic itself is no doubt both interesting and challenging, given that the same sentence spoken in different tones conveys different emotion; for example, if a job is not done, one can say “It’s not done” in different tones conveying anger, surprise, or even humor (I leave it to the reader to try this and confirm!).

The book is organized into eight chapters. Chapter 1 introduces the properties of a speech signal. It briefly discusses different types of emotion (for example, anger, sadness, happiness, fear, and so on) and their characteristics in a speech signal and how they can be captured in a computer. The application areas of speech emotion recognition are listed.

Chapters 2 and 3 cover emotion recognition using prosodic and spectral features, respectively. The prosodic features include mainly energy, zero crossing rate, and pitch, while the spectral features include mainly mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC).

Some other topics covered in this brief research monogram are feature fusion techniques (chapter 4), emotional speech corpora (chapter 5), classification models (chapter 6), and a “comparative analysis of classifiers in emotion recognition” (chapter 7). The last chapter (8) is reserved for the summary and conclusion. The essence of the monogram is that fusion of certain acoustical features greatly increases the performance of the emotion recognizing model.

On the plus side, the monogram looks concise and interesting and should be of interest to postgraduates and researchers in speech processing. On the minus side, the MATLAB codes that implement the acoustical features could have been provided through some web resource. The difference between speech and music signals should have been mentioned at the outset.

Although speech and music signals are both continuous functions of time, the latter, unlike the former, are not continuously varying. They remain fixed for the duration of a musical note and then jump to the pitch of the next note [1]. The term “fixed” here does not mean constant, however, as pitch is never steady, but the fundamental frequencies remain fairly close for the duration of a note. There are other differences also; for example, the rhythm in speech is not as regular as that in music. Also, I noticed a few grammatical errors, for example, “anger and happy emotions are in the same arousal state but they are differ in affect” (page 5). Finally, I would strongly encourage the authors to work on musical emotion, too, which is far more complex than human emotion. Let us not forget that even great writers, whose novels, stories, and other works do make us emotional, which implies they were quite capable of conveying emotion through words, did not take the risk of being music writers or critics.