Emotions are very subjective and scholars find it difficult to classify the range of emotions displayed by humans (and animals). Not only do we show emotions but equally well detect emotions of others. This production and detection of emotions happiness, sadness, joyousness, anger, disgust, fear, and surprise, helps us to function intelligently in this world. These emotions are articulated through facial expressions, tone of voice, gestures, and choice of words. Developments in key branches of social psychology, AI and statistical machine learning have led to algorithms for analysing face, and for analysing voice, in real-time and then estimating emotional state of a person. This is a rapidly developing product line for use in video phones to check your caller’s emotional state, in marketing of goods and services for detecting customer/consumer emotions and so on.
In this project you will learn about the emotion detection algorithms and implement these to build a computer system that has emotional intelligence enabling a machine to detect facial emotions, voice emotions, and emotions expressed through head movement. You will be working on the cutting edge of AI/ML as emotional intelligence is an essential human trait and it is very challenging to endow a machine with this trait
See one on my papers (i) Vogel, C., & Ahmad, K. (2023). Agreement and disagreement between major emotion recognition systems. Knowledge-Based Systems, 276, 110759. (https://www.sciencedirect.com/science/article/pii/S0950705123005099) (ii) Datta, D., Jiang, W., Vogel, C., & Ahmad, K. (2023, February). Speech Emotion Recognition Systems: A Cross-Language, Inter-racial, and Cross-Gender Comparison. In Future of Information and Communication Conference (pp. 375-390). Cham: Springer Nature Switzerland. (https://link.springer.com/chapter/10.1007/978-3-031-28076-4_28)