Humans typically express their emotions verbally or nonverbally as a response to an outside event. Several modalities, including text, audio, body motions, facial expressions, and physiological signs, can be used to conduct emotion recognition.

The field of human emotion recognition is dynamic because it has many uses in human-computer interaction. The goal of the authors' work is to develop intelligent and adaptive learning algorithms for the recognition of human emotions from physiological signals, micro-expressions, and facial expressions.

Using both posed and spontaneous facial expressions, precise and efficient deep learning models for the classification of human emotions have been described. For precise computational techniques, these models take advantage of the discrete wavelet transform and the self-attention mechanism.

This book also demonstrates the widespread acceptance of transformer models in language processing tasks because of their exceptional performance. Hence, human emotion recognition through micro-expressions has been achieved in this work by utilizing a modified version of the existing vision transformer.

Furthermore, using physiological inputs, specifically electroencephalograms (EEGs), a deep learning model for human emotion identification has been developed. This book primarily aims to accomplish two things: (i) to identify emotions from physiological patterns and facial expressions; and (ii) to develop deep learning frameworks that are intelligent and adaptive and solve the challenges associated with human emotion recognition.

. .