通过偏好对齐引导跨模态表示
Guiding Cross-Modal Representations with MLLM Priors via Preference Alignment
AI总结：
MAPLE通过利用多模态大语言模型的细粒度对齐先验，引导跨模态表示学习，提升细粒度检索性能。
摘要：
尽管对比语言-图像预训练（CLIP）在跨模态检索方面表现出色，但其特征空间中仍存在显著的模态差距。有趣的是，我们发现现成的多模态大语言模型（MLLM）表现出强大的内在模态对齐特性。虽然最近的基于MLLM的检索器采用了统一架构部分缓解了这一差距，但其依赖于粗粒度的模态对齐机制从根本上限制了其潜力。在本工作中，我们引入了MAPLE（用于嵌入的模态对齐偏好学习），这是一种新的框架，利用MLLM中固有的细粒度对齐先验来引导跨模态表示学习。MAPLE将学习过程建模为强化学习，包含两个关键组成部分：（1）使用现成的MLLM自动构建偏好数据，（2）一种新的相对偏好对齐（RPA）损失，将直接偏好优化（DPO）适应到嵌入学习设置中。实验结果表明，我们的偏好引导对齐在细粒度跨模态检索中实现了显著提升，凸显了其在处理微妙语义差异方面的有效性。
英文摘要：
Despite Contrastive Language-Image Pretraining (CLIP)'s remarkable capability to retrieve content across modalities, a substantial modality gap persists in its feature space. Intriguingly, we discover that off-the-shelf MLLMs (Multimodal Large Language Models) demonstrate powerful inherent modality alignment properties. While recent MLLM-based retrievers with unified architectures partially mitigate this gap, their reliance on coarse modality alignment mechanisms fundamentally limits their potential. In this work, We introduce MAPLE (Modality-Aligned Preference Learning for Embeddings), a novel framework that leverages the fine grained alignment priors inherent in MLLM to guide cross modal representation learning. MAPLE formulates the learning process as reinforcement learning with two key components: (1) Automatic preference data construction using off-the-shelf MLLM, and (2) a new Relative Preference Alignment (RPA) loss, which adapts Direct Preference Optimization (DPO) to the embedding learning setting. Experimental results show that our preference-guided alignment achieves substantial gains in fine-grained cross-modal retrieval, underscoring its effectiveness in handling nuanced semantic distinctions.
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。
