Published Paper Details:

Human expressions are vital to state their emotions. Facial expression recognition has always been a challenging task in recent times and also proved that it can be ef*ciently used in many areas. According to a recent survey, people usually try hiding their emotions at times when they are to be monitored. Almost all the decisions we take are driven by emotions. Marketing research has also proven that predicting sentiments correctly can be a huge source of growth for businesses and other sectors. It is very crucial for social interaction and has gained extensive attention from researchers for their research in various range of methods. The Convolution neural network as a powerful image processing and artificial intelligence network make use of deep learning to deal with both descriptive and analytical tasks using different elements and visions of smart machines to deploy the image and video recognition. Emotion recognition monitors the procedures of various emotional aspects of a human and displays the particular human emotion using the various concepts of Convolution neural networks. We have added many techniques of CNN to the FER-2013 data set and algorithms like Haar Cascade for facial recognition and frame development, and Image pre-processing techniques where the data set has been trained with various weights. The FER-2013 data set was created by gathering the results of a Google image search of each emotion and synonyms of the emotions. The data set has a total of 5,876 labeled images of 123 individuals. Out of these images, we used 4,113 images for training, 881 for dev, and 881 for the test. The main aim is to detect and recognize the facial expressions of students during their presence in any project review. Keeping in mind all the various methods to recognize facial emotions, a very important aspect is that using gray scaled images always makes the work easier without interrupting the quality and accuracy of the model.