A Sub-Layered Hierarchical Pyramidal Neural Architecture for Facial Expression Recognition

TL;DR

This paper proposes a new pyramidal neural architecture with a connectivity scheme that enhances feature learning, offering comparable facial expression recognition performance to convolutional networks but with fewer parameters and better robustness for low-resolution images.

Contribution

Introduces a connectivity scheme for pyramidal neural architectures that improves learning capacity and efficiency for facial expression recognition in resource-constrained settings.

Findings

Performs as well as convolutional networks on facial expression recognition

Uses fewer trainable parameters than traditional CNNs

More robust to low-resolution face images

Abstract

In domains where computational resources and labeled data are limited, such as in robotics, deep networks with millions of weights might not be the optimal solution. In this paper, we introduce a connectivity scheme for pyramidal architectures to increase their capacity for learning features. Experiments on facial expression recognition of unseen people demonstrate that our approach is a potential candidate for applications with restricted resources, due to good generalization performance and low computational cost. We show that our approach generalizes as well as convolutional architectures in this task but uses fewer trainable parameters and is more robust for low-resolution faces.