Face Behavior a la carte: Expressions, Affect and Action Units in a Single Network

TL;DR

This paper introduces FaceBehaviorNet, a multi-task neural network trained on large-scale datasets to jointly recognize facial expressions, emotions, and action units, outperforming single-task models and enabling zero- and few-shot learning for related tasks.

Contribution

It presents the first large-scale joint training of facial behavior tasks in a single network, demonstrating improved performance and versatile feature learning.

Findings

Joint training outperforms single-task models.

Proposed coupling strategies enhance training effectiveness.

Features learned enable zero- and few-shot recognition of new tasks.

Abstract

Automatic facial behavior analysis has a long history of studies in the intersection of computer vision, physiology and psychology. However it is only recently, with the collection of large-scale datasets and powerful machine learning methods such as deep neural networks, that automatic facial behavior analysis started to thrive. Three of its iconic tasks are automatic recognition of basic expressions (e.g. happy, sad, surprised), estimation of continuous emotions (e.g., valence and arousal), and detection of facial action units (activations of e.g. upper/inner eyebrows, nose wrinkles). Up until now these tasks have been mostly studied independently collecting a dataset for the task. We present the first and the largest study of all facial behaviour tasks learned jointly in a single multi-task, multi-domain and multi-label network, which we call FaceBehaviorNet. For this we utilize all publicly available datasets in the community (around 5M images) that study facial behaviour tasks in-the-wild. We demonstrate that training jointly an end-to-end network for all tasks has consistently better performance than training each of the single-task networks. Furthermore, we propose two simple strategies for coupling the tasks during training, co-annotation and distribution matching, and show the advantages of this approach. Finally we show that FaceBehaviorNet has learned features that encapsulate all aspects of facial behaviour, and can be successfully applied to perform tasks (compound emotion recognition) beyond the ones that it has been trained in a zero- and few-shot learning setting.