Recognize facial expression using CNN, SVM and CNN-SVM models.

Facial expression recognition from images have attracted much research interest within the field of Deep Learning. Despite much efforts made in developing various methods for expression recognition, traditional machine learning approaches lack generalizability when applied to unseen images.

There are many use cases in real life which could lead to advances in psychology and sociology, and an increased understanding of decision making in Robotics and consumer preferences.

The traditional solution uses the approach that emotions depends on the locations of facial landmarks and that emotions are dependent on the relationship between landmarks. Our first approach will be to use an SVM because it has the ability to capture complex relationships in both regression and classification problems[3]. We will also analyse this problem using convolutional neural networks (CNNs).

1. Justin Shiao
2. Anchit Agarwal
3. Anurag Solanki
4. Abhishek Kumar