Open AccessFacial Expression Recognition using CNN
Author(s) -
S. Preethi,
P. G. Mary Sobha,
P. Rajalakshmi Kamalini,
K. Gowri Raghavendra Narayan
Publication year - 2020
Publication title -
international journal of scientific research in computer science, engineering and information technology
Language(s) - English
Resource type - Journals
ISSN - 2456-3307
DOI - 10.32628/cseit206248
Subject(s) - disgust , surprise , convolutional neural network , computer science , facial expression , confusion , artificial intelligence , confusion matrix , expression (computer science) , feeling , facial recognition system , speech recognition , face (sociological concept) , pattern recognition (psychology) , psychology , communication , anger , social psychology , social science , sociology , psychoanalysis , programming language
People have consistently been able to perceive and recognize faces and their feelings. Presently PCs can do likewise. We propose a model which recognizes human faces and classifies the emotion on the face as happy, angry, sad, neutral, surprise, disgust or fear. It is developed utilizing a convolutional neural network(CNN) and involves various stages. All these are carried out using a dataset available on the Kaggle repository named fer2013. Precision and execution of the neural system can be assessed utilizing a confusion matrix. We applied cross-approval to decide the ideal hyper-parameters and assessed the presentation of the created models by looking at their training histories.