MGMR-Net: Mamba-Guided Multimodal Reconstruction and Fusion Network for Sentiment Analysis with Incomplete Modalities
Electronics · 01.08.2025 · 14 (15) · 3088
DOI: 10.3390/electronics14153088 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 01.08.2025. Verfasst von Chengcheng Yang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangdong Basic and Applied Basic Research Foundation, Science and Technology Program of Guangzhou, Heyuan Social Science and Agriculture Project.
Abstract (Originalsprache)
Multimodal sentiment analysis (MSA) faces key challenges such as incomplete modality inputs, long-range temporal dependencies, and suboptimal fusion strategies. To address these, we propose MGMR-Net, a Mamba-guided multimodal reconstruction and fusion network that integrates modality-aware reconstruction with text-centric fusion within an efficient state-space modeling framework. MGMR-Net consists of two core components: the Mamba-collaborative fusion module, which utilizes a two-stage selective state-space mechanism for fine-grained cross-modal alignment and hierarchical temporal integration, and the Mamba-enhanced reconstruction module, which employs continuous-time recurrence and dynamic gating to accurately recover corrupted or missing modality features. The entire network is jointly optimized via a unified multi-task loss, enabling simultaneous learning of discriminative features for sentiment prediction and reconstructive features for modality recovery. Extensive experiments on CMU-MOSI, CMU-MOSEI, and CH-SIMS datasets demonstrate that MGMR-Net consistently outperforms several baseline methods under both complete and missing modality settings, achieving superior accuracy, robustness, and generalization.
Förderung
- Guangdong Basic and Applied Basic Research Foundation
- Science and Technology Program of Guangzhou
- Heyuan Social Science and Agriculture Project
- Science and Technology Planning Project of Yunfu
- Key Construction Discipline Research Ability Enhancement Project of Guangdong Province
- Guangdong Province Science and Technology Innovation Strategy Special Fund
Zitieren
APA
Yang, C., Liang, Z., Liu, T., Hu, Z., & Yan, D. (2025). MGMR-Net: Mamba-Guided Multimodal Reconstruction and Fusion Network for Sentiment Analysis with Incomplete Modalities. Electronics, 14(15), 3088. https://doi.org/10.3390/electronics14153088
Vancouver
Yang C, Liang Z, Liu T, Hu Z, Yan D. MGMR-Net: Mamba-Guided Multimodal Reconstruction and Fusion Network for Sentiment Analysis with Incomplete Modalities. Electronics. 2025;14(15):3088. doi:10.3390/electronics14153088
BibTeX
@article{Yang2025,
author = {Yang, Chengcheng and Liang, Zhiyao and Liu, Tonglai and Hu, Zeng and Yan, Dashun},
title = {MGMR-Net: Mamba-Guided Multimodal Reconstruction and Fusion Network for Sentiment Analysis with Incomplete Modalities},
journal = {Electronics},
year = {2025},
volume = {14},
number = {15},
pages = {3088},
doi = {10.3390/electronics14153088}
}
EUVIMED