Generative Adversarial Stacked Autoencoders for Facial Pose Normalization and Emotion Recognition

TL;DR

This paper introduces a novel Generative Adversarial Stacked Autoencoder that normalizes facial pose and improves emotion recognition accuracy across diverse datasets using innovative convolutional layers and a new training algorithm.

Contribution

It presents a new autoencoder architecture with a specialized convolutional layer and a novel training algorithm for robust facial pose normalization and emotion recognition.

Findings

Achieved state-of-the-art performance on multiple facial emotion datasets.

Effectively normalizes facial pose up to ±60 degrees.

Demonstrates efficiency and robustness in wild conditions.

Abstract

In this work, we propose a novel Generative Adversarial Stacked Autoencoder that learns to map facial expressions, with up to plus or minus 60 degrees, to an illumination invariant facial representation of 0 degrees. We accomplish this by using a novel convolutional layer that exploits both local and global spatial information, and a convolutional layer with a reduced number of parameters that exploits facial symmetry. Furthermore, we introduce a generative adversarial gradual greedy layer-wise learning algorithm designed to train Adversarial Autoencoders in an efficient and incremental manner. We demonstrate the efficiency of our method and report state-of-the-art performance on several facial emotion recognition corpora, including one collected in the wild.