Simple CNN model for FER2013 dataset with 64.78 accuracy on test data. The FER2013[1], was a challenge proposed on Kaggle which was won by the team reaching the test accuracy of 75.2%. The data set consists various 48x48 images partitioned into training, validation and testing data. The images are labelled for emotions and the labels are as follows:

• Angry
• Disgust
• Fear
• Happy
• Sad
• Surprise
• Neutral

Human accuracy on the dataset is 65.5%[2] and a simple CNN model can be trained to achieve that accuracy.

1. I. J. Goodfellow, D. Erhan, P. L. Carrier, A. Courville, M. Mirza, B. Hamner, W. Cukierski, Y. Tang, D. Thaler, D.-H. Lee, Y. Zhou, C. Ramaiah, F. Feng, R. Li, X. Wang, D. Athanasakis, J. Shawe-Taylor, M. Milakov, J. Park, R. Ionescu, M. Popescu, C. Grozea, J. Bergstra, J. Xie, L. Romaszko, B. Xu, Z. Chuang, and Y. Bengio. Challenges in representation learning: A report on three machine learning contests. Neural Networks, 64:59--63, 2015. Special Issue on "Deep Learning of Representations"
2. Christopher Pramerdorfer, Martin Kampel, Facial Expression Recognition using Convolutional Neural Networks: State of the Art, https://arxiv.org/pdf/1612.02903.pdf