CNN based Mood Detection using Facial Expression

Life nowadays is generally much easier than in the past due to the advancements in the technology. Technology even helps us in identifying one's emotion from an image or video without any human intervention. Understanding the emotions of individuals is important in order to know student's feelings after a lecture, which shows their understanding level, employee's emotions after late working hours and many more. Humans convey a lot of information through their facial expressions in addition to their words. Thus, accurate expression recognition has become crucial for any modern human-computer interface. Here, this research study proposes a novel technique for recognising facial expressions based on CNN architecture. The primary driving force behind this study is to increase the deep learning model's accuracy as it might have an effect on future work. The model achie ve d the accuracy of 89.91% accuracy on FER-2013 and 96.01% on CK+.