Ada-DF: An Adaptive Label Distribution Fusion Network For Facial Expression Recognition

TL;DR

This paper introduces Ada-DF, a dual-branch adaptive fusion network for facial expression recognition that effectively handles annotation ambiguity by leveraging label distribution learning, demonstrating superior performance on multiple datasets.

Contribution

The paper proposes a novel dual-branch adaptive distribution fusion framework for FER that adaptively combines label distributions to improve recognition accuracy.

Findings

Ada-DF outperforms state-of-the-art methods on RAF-DB, AffectNet, and SFEW datasets.

The auxiliary branch effectively captures label distributions for ambiguous annotations.

Adaptive fusion improves robustness and accuracy in FER tasks.

Abstract

Facial expression recognition (FER) plays a significant role in our daily life. However, annotation ambiguity in the datasets could greatly hinder the performance. In this paper, we address FER task via label distribution learning paradigm, and develop a dual-branch Adaptive Distribution Fusion (Ada-DF) framework. One auxiliary branch is constructed to obtain the label distributions of samples. The class distributions of emotions are then computed through the label distributions of each emotion. Finally, those two distributions are adaptively fused according to the attention weights to train the target branch. Extensive experiments are conducted on three real-world datasets, RAF-DB, AffectNet and SFEW, where our Ada-DF shows advantages over the state-of-the-art works.