Grounding Emotion Recognition with Visual Prototypes: VEGA -- Revisiting CLIP in MERC

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Hu, Guanyu, Kollias, Dimitrios, Yang, Xinyu
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912536225382400
author Hu, Guanyu
Kollias, Dimitrios
Yang, Xinyu
author_facet Hu, Guanyu
Kollias, Dimitrios
Yang, Xinyu
contents Multimodal Emotion Recognition in Conversations remains a challenging task due to the complex interplay of textual, acoustic and visual signals. While recent models have improved performance via advanced fusion strategies, they often lack psychologically meaningful priors to guide multimodal alignment. In this paper, we revisit the use of CLIP and propose a novel Visual Emotion Guided Anchoring (VEGA) mechanism that introduces class-level visual semantics into the fusion and classification process. Distinct from prior work that primarily utilizes CLIP's textual encoder, our approach leverages its image encoder to construct emotion-specific visual anchors based on facial exemplars. These anchors guide unimodal and multimodal features toward a perceptually grounded and psychologically aligned representation space, drawing inspiration from cognitive theories (prototypical emotion categories and multisensory integration). A stochastic anchor sampling strategy further enhances robustness by balancing semantic stability and intra-class diversity. Integrated into a dual-branch architecture with self-distillation, our VEGA-augmented model achieves sota performance on IEMOCAP and MELD. Code is available at: https://github.com/dkollias/VEGA.
format Preprint
id arxiv_https___arxiv_org_abs_2508_06564
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Grounding Emotion Recognition with Visual Prototypes: VEGA -- Revisiting CLIP in MERC
Hu, Guanyu
Kollias, Dimitrios
Yang, Xinyu
Computer Vision and Pattern Recognition
Multimodal Emotion Recognition in Conversations remains a challenging task due to the complex interplay of textual, acoustic and visual signals. While recent models have improved performance via advanced fusion strategies, they often lack psychologically meaningful priors to guide multimodal alignment. In this paper, we revisit the use of CLIP and propose a novel Visual Emotion Guided Anchoring (VEGA) mechanism that introduces class-level visual semantics into the fusion and classification process. Distinct from prior work that primarily utilizes CLIP's textual encoder, our approach leverages its image encoder to construct emotion-specific visual anchors based on facial exemplars. These anchors guide unimodal and multimodal features toward a perceptually grounded and psychologically aligned representation space, drawing inspiration from cognitive theories (prototypical emotion categories and multisensory integration). A stochastic anchor sampling strategy further enhances robustness by balancing semantic stability and intra-class diversity. Integrated into a dual-branch architecture with self-distillation, our VEGA-augmented model achieves sota performance on IEMOCAP and MELD. Code is available at: https://github.com/dkollias/VEGA.
title Grounding Emotion Recognition with Visual Prototypes: VEGA -- Revisiting CLIP in MERC
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.06564