The FaceChannelS: Strike of the Sequences for the AffWild 2 Challenge

TL;DR

This paper benchmarks various versions of the FaceChannel neural network for predicting affective information from facial expressions on the new AffWild2 dataset, highlighting the model's adaptability and performance.

Contribution

It introduces a new benchmarking study of FaceChannel models on the AffWild2 dataset, demonstrating their effectiveness in affect prediction tasks.

Findings

FaceChannel models achieve competitive results on AffWild2.

The models show good adaptability to new affective datasets.

Benchmarking provides insights into model performance and limitations.

Abstract

Predicting affective information from human faces became a popular task for most of the machine learning community in the past years. The development of immense and dense deep neural networks was backed by the availability of numerous labeled datasets. These models, most of the time, present state-of-the-art results in such benchmarks, but are very difficult to adapt to other scenarios. In this paper, we present one more chapter of benchmarking different versions of the FaceChannel neural network: we demonstrate how our little model can predict affective information from the facial expression on the novel AffWild2 dataset.