Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2603.18462 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915874310455296 |
|---|---|
| author | Li, Yan Xing, Yifei Lan, Xiangyuan Li, Xin Chen, Haifeng Jiang, Dongmei |
| author_facet | Li, Yan Xing, Yifei Lan, Xiangyuan Li, Xin Chen, Haifeng Jiang, Dongmei |
| contents | In the era of large-scale pre-trained models, effectively adapting general knowledge to specific affective computing tasks remains a challenge, particularly regarding computational efficiency and multimodal heterogeneity. While Transformer-based methods have excelled at modeling inter-modal dependencies, their quadratic computational complexity limits their use with long-sequence data. Mamba-based models have emerged as a computationally efficient alternative; however, their inherent sequential scanning mechanism struggles to capture the global, non-sequential relationships that are crucial for effective cross-modal alignment. To address these limitations, we propose \textbf{AlignMamba-2}, an effective and efficient framework for multimodal fusion and sentiment analysis. Our approach introduces a dual alignment strategy that regularizes the model using both Optimal Transport distance and Maximum Mean Discrepancy, promoting geometric and statistical consistency between modalities without incurring any inference-time overhead. More importantly, we design a Modality-Aware Mamba layer, which employs a Mixture-of-Experts architecture with modality-specific and modality-shared experts to explicitly handle data heterogeneity during the fusion process. Extensive experiments on four challenging benchmarks, including dynamic time-series (on the CMU-MOSI and CMU-MOSEI datasets) and static image-related tasks (on the NYU-Depth V2 and MVSA-Single datasets), demonstrate that AlignMamba-2 establishes a new state-of-the-art in both effectiveness and efficiency across diverse pattern recognition tasks, ranging from dynamic time-series analysis to static image-text classification. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_18462 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AlignMamba-2: Enhancing Multimodal Fusion and Sentiment Analysis with Modality-Aware Mamba Li, Yan Xing, Yifei Lan, Xiangyuan Li, Xin Chen, Haifeng Jiang, Dongmei Artificial Intelligence In the era of large-scale pre-trained models, effectively adapting general knowledge to specific affective computing tasks remains a challenge, particularly regarding computational efficiency and multimodal heterogeneity. While Transformer-based methods have excelled at modeling inter-modal dependencies, their quadratic computational complexity limits their use with long-sequence data. Mamba-based models have emerged as a computationally efficient alternative; however, their inherent sequential scanning mechanism struggles to capture the global, non-sequential relationships that are crucial for effective cross-modal alignment. To address these limitations, we propose \textbf{AlignMamba-2}, an effective and efficient framework for multimodal fusion and sentiment analysis. Our approach introduces a dual alignment strategy that regularizes the model using both Optimal Transport distance and Maximum Mean Discrepancy, promoting geometric and statistical consistency between modalities without incurring any inference-time overhead. More importantly, we design a Modality-Aware Mamba layer, which employs a Mixture-of-Experts architecture with modality-specific and modality-shared experts to explicitly handle data heterogeneity during the fusion process. Extensive experiments on four challenging benchmarks, including dynamic time-series (on the CMU-MOSI and CMU-MOSEI datasets) and static image-related tasks (on the NYU-Depth V2 and MVSA-Single datasets), demonstrate that AlignMamba-2 establishes a new state-of-the-art in both effectiveness and efficiency across diverse pattern recognition tasks, ranging from dynamic time-series analysis to static image-text classification. |
| title | AlignMamba-2: Enhancing Multimodal Fusion and Sentiment Analysis with Modality-Aware Mamba |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.18462 |