Dual Stream Independence Decoupling for True Emotion Recognition under Masked Expressions

TL;DR

This paper introduces a new apexframe-based paradigm and a dual stream independence decoupling framework to improve true emotion recognition from masked expressions, effectively separating true and disguised emotion features.

Contribution

It proposes a novel apexframe-based paradigm and a dual stream decoupling framework with a specialized loss group for better true emotion recognition under masks.

Findings

Apexframe-based paradigm is more challenging but effective.

Decoupling framework improves recognition accuracy.

Enhanced independence of true and disguised emotion features.

Abstract

Recongnizing true emotions from masked expressions is extremely challenging due to deliberate concealment. Existing paradigms recognize true emotions from masked-expression clips that contain onsetframes just starting to disguise. However, this paradigm may not reflect the actual disguised state, as the onsetframe leaks the true emotional information without reaching a stable disguise state. Thus, this paper introduces a novel apexframe-based paradigm that classifies true emotions from the apexframe with a stable disguised state. Furthermore, this paper proposes a novel dual stream independence decoupling framework that decouples true and disguised emotion features, avoiding the interference of disguised emotions on true emotions. For efficient decoupling, we design a decoupling loss group, comprising two classification losses that learn true emotion and disguised expression features, respectively, and a Hilbert-Schmidt Independence loss that enhances the independence of two features. Experiments demonstrate that the apexframe-based paradigm is challenging, and the proposed decouple framework improves recogntion performances.