arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2601.09746cs.MA
收藏

多智能体协作学习用于应对异常概念下的鲁棒视觉语言对齐

Multi-Agent Cooperative Learning for Robust Vision-Language Alignment under OOD Concepts

Philip Xu

AI总结:

本文提出多智能体协作学习框架,通过结构化信息传递缓解模态不平衡,提升视觉语言模型在异常概念下的对齐性能,实验表明在少样本和零样本设置中精度提升1-5%。

摘要:

本文介绍了一种新颖的多智能体协作学习(MACL)框架,旨在解决视觉语言模型在处理分布外(OOD)概念时跨模态对齐崩溃的问题。四个核心智能体,包括图像、文本、名称和协调智能体,通过结构化信息传递协同缓解模态不平衡。所提框架实现了多智能体特征空间名称学习,融合了上下文交换增强的少样本学习算法,并采用自适应动态平衡机制调节智能体间的贡献。在VISTA-Beyond数据集上的实验表明,MACL在少样本和零样本设置中显著提升了性能,实现了在多种视觉领域中的1-5%精度提升。

英文摘要:

This paper introduces a novel Multi-Agent Cooperative Learning (MACL) framework to address cross-modal alignment collapse in vision-language models when handling out-of-distribution (OOD) concepts. Four core agents, including image, text, name, and coordination agents, collaboratively mitigate modality imbalance through structured message passing. The proposed framework enables multi-agent feature space name learning, incorporates a context exchange enhanced few-shot learning algorithm, and adopts an adaptive dynamic balancing mechanism to regulate inter-agent contributions. Experiments on the VISTA-Beyond dataset demonstrate that MACL significantly improves performance in both few-shot and zero-shot settings, achieving 1-5% precision gains across diverse visual domains.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑