JJEM: Special Edition 2 (December 2024)
JJEM: Special Edition 2 (December 2024)
2024-12-08
Facial Expression based sentimental analysis using CNN.
M Subrahmanya Adiga, Adarsh M J
A developing area of affective computing is the study of facial changes employing CNN to artificially emulate the capability of interpreting people’s feelings. The FER-13 dataset, which consists of a variety of face photos labelled with the following seven fundamental emotions: Having shown interest toward this work, seven basic emotions are employed throughout the study, which include anger, contempt, fear, happiness, sorrow, surprise, and the last one being a neutral emotion. Specifically, the major goal is to improve the effectiveness of sentiment analysis by applying advanced CNN architectures. The main advantages of FER-13 dataset are acknowledged and ensure methodological reliability of the work, since the functionality of the created model will be tested utilizing this extensive source. In this way, the CNN method is analysed to differentiate the attitudes and, thus, has a more profound understanding of persons’ emotions, which is laudable when compared to a vast linguistic and cultural difference that is excluded in large data sets. Some of the recommendations are including pre-processing of the data to normalize and enrich the images. It is followed by the training of the CNN model employing optimum hyperparameters with the purpose of achieving a high classification accuracy. By carrying out the various tests and cross validation techniques, one is in a position to validate the model. The findings showed considerable enhancements compared to previous methods. accuracy of sentiment classification that can highlight the potential of CNN-oriented approaches in relatable real-world applications including mental health monitoring, social media, and robots, and humans’ interpersonal communication.
CNN, Neural Network, Facial Expression, dataset, AI.