Affective Behaviour Analysis via Integrating Multi-Modal Knowledge

TL;DR

This paper introduces a multi-modal emotion analysis method utilizing transformer-based fusion, masked autoencoders for facial features, and scene-specific training, achieving superior performance across five affective behavior tasks in natural environments.

Contribution

The paper presents a novel multi-modal emotion analysis framework combining transformer fusion, masked autoencoder visual features, and scene-aware training for improved affective behavior recognition.

Findings

Demonstrated superior performance in five affective behavior tasks

Effective multi-modal feature integration using transformer modules

Enhanced facial feature extraction with fine-tuned Masked-Auto Encoder

Abstract

Affective Behavior Analysis aims to facilitate technology emotionally smart, creating a world where devices can understand and react to our emotions as humans do. To comprehensively evaluate the authenticity and applicability of emotional behavior analysis techniques in natural environments, the 6th competition on Affective Behavior Analysis in-the-wild (ABAW) utilizes the Aff-Wild2, Hume-Vidmimic2, and C-EXPR-DB datasets to set up five competitive tracks, i.e., Valence-Arousal (VA) Estimation, Expression (EXPR) Recognition, Action Unit (AU) Detection, Compound Expression (CE) Recognition, and Emotional Mimicry Intensity (EMI) Estimation. In this paper, we present our method designs for the five tasks. Specifically, our design mainly includes three aspects: 1) Utilizing a transformer-based feature fusion module to fully integrate emotional information provided by audio signals, visual images, and transcripts, offering high-quality expression features for the downstream tasks. 2) To achieve high-quality facial feature representations, we employ Masked-Auto Encoder as the visual features extraction model and fine-tune it with our facial dataset. 3) Considering the complexity of the video collection scenes, we conduct a more detailed dataset division based on scene characteristics and train the classifier for each scene. Extensive experiments demonstrate the superiority of our designs.