Research Area:  Machine Learning
Humans express their innate emotions and conditions through various channels such as body language and facial expressions. Facial expressions are the most direct and meaningful channel of non-verbal communication which forms a universal language of emotions that can instantly express a wide range of human emotional states, feelings, attitudes and thus assists in various cognitive tasks. The accurate analysis and interpretation of the emotional content of human facial expressions is essential for the deeper understanding of human behavior. In this work, we study the field of Facial Expression Recognition in-depth, and perform Emotion Recognition on 2D images. The task is supervised classification of data, for which we implement a wide range of Deep Learning models in order to recognize the existent emotion. Furthermore, an additional topic showcased in this work is Attention-based models. A concrete experimental study was designed and conducted and the results are quite interesting.
Keywords:  
Facial Expression Recognition
Deep Learning
Machine Learning
Author(s) Name:  Spyridon Kardakis, Isidoros Perikos, Foteini Grivokostopoulou & Michael Paraskevas
Journal name:  
Conferrence name:  International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing
Publisher name:  Springer
DOI:  10.1007/978-3-030-92317-4_13
Volume Information:  
Paper Link:   https://link.springer.com/chapter/10.1007/978-3-030-92317-4_13