Two-tier Emergent Self-Organizing (TtEsom) approach of understanding emotions

This paper extends the previous work on emotion mapping that attempts to emulate human brain reference model. Most emotion recognition system analyzes facial expression through supervised learning whereas this work adopts unsupervised learning. The system first locates the human face in an image, and then identifies the localized face emotion. The proposed method uses features obtained using Gabor wavelets, undergoes features selection through the use of a derivation of Support Vector Machine. This work adopted a connectionist model, called Two-tier Emergent Self-Organizing Map (TtEsom) to analyse the emotion. The result shows improvement over the previous work and comparable result with supervised learning approach.