Bayesian Convolutional Neural Networks for Seven Basic Facial Expression Classifications

TL;DR

This paper introduces a Bayesian convolutional neural network based on ResNet18 for facial expression classification, achieving improved accuracy by novel objective functions and training schemes, tested on FER2013 dataset.

Contribution

The paper proposes a new Bayesian neural network framework with a unique objective function and training method, focusing on last convolutional layer parameters for better facial expression classification.

Findings

Achieved 71.5% accuracy on PublicTestSet

Achieved 73.1% accuracy on PrivateTestSet

Outperformed traditional Bayesian neural networks in accuracy

Abstract

The seven basic facial expression classifications are a basic way to express complex human emotions and are an important part of artificial intelligence research. Based on the traditional Bayesian neural network framework, the ResNet18_BNN network constructed in this paper has been improved in the following three aspects: (1) A new objective function is proposed, which is composed of the KL loss of uncertain parameters and the intersection of specific parameters. Entropy loss composition. (2) Aiming at a special objective function, a training scheme for alternately updating these two parameters is proposed. (3) Only model the parameters of the last convolution group. Through testing on the FER2013 test set, we achieved 71.5% and 73.1% accuracy in PublicTestSet and PrivateTestSet, respectively. Compared with traditional Bayesian neural networks, our method brings the highest classification accuracy gain.