Main Article Content
In this paper, using two distinct media, video and text, the model will undertake emotion analysis in order to classify participants into various categories, which will result in the identification and categorization of different emotions that are being displayed by the participants. In this paper, Convolutional Neural Network, in assistance with OpenCV, Keras, Tenorflow and Numpy, has been used to train the emotion classifying model so that it could distinguish different emotions. The model has been trained and validated with over thirty-five thousand images to increase accuracy. Pre-processed images are supplied from the database to the CNN model and it undergoes all the layers to make sure the model can accurately identify the patterns of the different emotions and finally identify it from a emotion dictionary using prediction and matching the pattern matching with the highest percentage with the respective emotion which are: happy, angry, disgust, neutral, sad, fear and surprise.