Recently, automatic facial emotion recognition (FER), which aims to analyze and understand human emotions using facial expressions, has become an increasingly active research topic in the domains of computer vision, artificial intelligence, pattern recognition, etc. This is because FER has many potential fields such as: e-learning, e-commerce, neuromarketing and healthcare systems. In the field of e-learning, it is well known that positive emotions, such as curiosity and joy, can lead students to engage more deeply in learning, while emotions such as boredom and frustration can lead students to resort to unproductive behaviors. FERs have been used to recognize these emotions which are used to develop student models that predict about the student learning. Results from these models (such as whether the student is engaged or unproductive) are then used to enhance the adaptive online learning environments. The project aims to, systematically evaluate the existing FER networks for analyzing the emotions of Indians, develop mechanisms to obtain ground truth information that can help validate results from FER and perform experiments to develop a labelled Indian dataset for emotion recognition.