ReINTEL: A Multimodal Data Challenge for Responsible Information Identification on Social Network Sites

Attentive Cross-modal Connections for Deep Multimodal Wearable-based Emotion Recognition

Classification of human emotions can play an essential role in the design and improvement of human-machine systems. While individual biological signals such as Electrocardiogram (ECG) and Electrodermal Activity (EDA) have been widely used for emotion recognition with machine learning methods, multimodal approaches generally fuse extracted features or final classification/regression results to boost performance. To enhance multimodal learning, we present a novel attentive cross-modal connection to share information between convolutional neural networks responsible for learning individual modali…

深度多模态子空间聚类网络

Deep Multimodal Subspace Clustering Networks

提出基于CNN的无监督多模态子空间聚类框架，包含多模态编码器、自表达层和解码器，通过空间融合和亲和融合策略，显著优于现有方法。

解决多模态歧义下性别化经济梗推理中的过度承诺问题

Addressing Overcommitment in the Reasoning of Gendered Economic Memes under Multimodal Ambiguity

本研究针对多模态歧义下性别化经济梗推理的认知过度承诺偏差，提出CGER-Net框架，在EconMeme-GE数据集上使歧义实例的性别过度承诺率降低最多44%，且79%的生成理由符合证据要求。

揭示不确定性：多模态大语言模型校准与性能的深度探究

Unveiling Uncertainty: A Deep Dive into Calibration and Performance of Multimodal Large Language Models

本文系统研究多模态大语言模型的校准问题，构建IDK数据集揭示其倾向作答而非承认不确定性的缺陷，并提出温度缩放与迭代提示优化技术以提升模型可靠性。

基于图注意力的多模态阿尔茨海默病相关基因优先级排序

Graph Attention Based Prioritization of Disease Responsible Genes from Multimodal Alzheimer's Network

NETRA通过多模态图变换器框架，利用注意力机制优先排序阿尔茨海默病相关基因，显著提升疾病通路富集分析效果。

通过可穿戴设备的边缘AI检测神经血管不稳定性：一种负责任的计算框架

Detecting Neurovascular Instability from Multimodal Physiological Signals Using Wearable-Compatible Edge AI: A Responsible Computational Framework

本文提出Melaguard框架，通过融合多种生理信号检测神经血管不稳定性，优于单一模态设备，通过三阶段验证提升检测性能。