Search papers, labs, and topics across Lattice.
This paper introduces a new paradigm for masked emotion recognition based on apex frames, which represent a stable disguised state, unlike prior work using onset frames. They propose a dual-stream architecture to decouple true and disguised emotion features using a novel decoupling loss group. Experiments demonstrate the challenge of the apexframe paradigm and the effectiveness of the proposed decoupling framework in improving recognition performance.
Recognizing emotions from deliberately disguised facial expressions is now possible thanks to a new method that focuses on the apex of the disguise, rather than its onset.
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.
