Federated Dialogue-Semantic Diffusion for Emotion Recognition under Incomplete Modalities

TL;DR

This paper introduces FedDISC, a federated learning framework for multimodal emotion recognition that effectively handles missing modalities by integrating diffusion models, dialogue context, and semantic consistency, outperforming existing methods.

Contribution

The paper presents a novel federated diffusion-based framework for missing-modality recovery in emotion recognition, incorporating dialogue context and semantic alignment, with a new alternating training strategy.

Findings

FedDISC outperforms existing methods on IEMOCAP, CMUMOSI, and CMUMOSEI datasets.

The framework effectively handles diverse missing modality patterns.

Semantic consistency is maintained through dialogue-aware diffusion models.

Abstract

Multimodal Emotion Recognition in Conversations (MERC) enhances emotional understanding through the fusion of multimodal signals. However, unpredictable modality absence in real-world scenarios significantly degrades the performance of existing methods. Conventional missing-modality recovery approaches, which depend on training with complete multimodal data, often suffer from semantic distortion under extreme data distributions, such as fixed-modality absence. To address this, we propose the Federated Dialogue-guided and Semantic-Consistent Diffusion (FedDISC) framework, pioneering the integration of federated learning into missing-modality recovery. By federated aggregation of modality-specific diffusion models trained on clients and broadcasting them to clients missing corresponding modalities, FedDISC overcomes single-client reliance on modality completeness. Additionally, the DISC-Diffusion module ensures consistency in context, speaker identity, and semantics between recovered and available modalities, using a Dialogue Graph Network to capture conversational dependencies and a Semantic Conditioning Network to enforce semantic alignment. We further introduce a novel Alternating Frozen Aggregation strategy, which cyclically freezes recovery and classifier modules to facilitate collaborative optimization. Extensive experiments on the IEMOCAP, CMUMOSI, and CMUMOSEI datasets demonstrate that FedDISC achieves superior emotion classification performance across diverse missing modality patterns, outperforming existing approaches.