VCU-Bridge: 通过语义桥梁实现层次化视觉隐喻理解
VCU-Bridge: Hierarchical Visual Connotation Understanding via Semantic Bridging
AI总结：
VCU-Bridge通过语义桥梁实现层次化视觉隐喻理解，构建了HVCU-Bench基准，并展示了提升MLLM能力的显著效果。
摘要：
尽管多模态大语言模型（MLLMs）在基准测试中表现出色，但其处理范式与人类整合视觉信息的能力存在差异。与人类自然地将细节和高层概念进行连接不同，模型倾向于将这些元素孤立处理。现有的评估协议往往将低层感知与高层推理割裂开来，忽视了它们的语义和因果依赖关系，导致非诊断性结果并掩盖了性能瓶颈。我们提出了VCU-Bridge框架，该框架实现了人类样式的视觉隐喻理解层次：多层次推理，从基础感知通过语义桥梁到抽象隐喻，具有从具体线索到抽象结论的明确证据到推理追溯。基于此框架，我们构建了HVCU-Bench，一个用于层次化视觉隐喻理解的基准，具有明确的层级诊断。全面的实验表明，随着推理向更高层次推进，性能呈现一致下降。我们进一步开发了一条由蒙特卡洛树搜索（MCTS）指导的指令微调数据生成管道，并展示了加强低层能力在更高层次上带来了可测量的提升。有趣的是，它不仅在HVCU-Bench上取得了改进，还在通用基准（平均+2.53%）上带来了好处，尤其是在MMStar上（+7.26%），证明了层次化思维模式的重要性和其在增强MLLM能力上的有效性。项目页面位于https://vcu-bridge.github.io。
英文摘要：
While Multimodal Large Language Models (MLLMs) excel on benchmarks, their processing paradigm differs from the human ability to integrate visual information. Unlike humans who naturally bridge details and high-level concepts, models tend to treat these elements in isolation. Prevailing evaluation protocols often decouple low-level perception from high-level reasoning, overlooking their semantic and causal dependencies, which yields non-diagnostic results and obscures performance bottlenecks. We present VCU-Bridge, a framework that operationalizes a human-like hierarchy of visual connotation understanding: multi-level reasoning that advances from foundational perception through semantic bridging to abstract connotation, with an explicit evidence-to-inference trace from concrete cues to abstract conclusions. Building on this framework, we construct HVCU-Bench, a benchmark for hierarchical visual connotation understanding with explicit, level-wise diagnostics. Comprehensive experiments demonstrate a consistent decline in performance as reasoning progresses to higher levels. We further develop a data generation pipeline for instruction tuning guided by Monte Carlo Tree Search (MCTS) and show that strengthening low-level capabilities yields measurable gains at higher levels. Interestingly, it not only improves on HVCU-Bench but also brings benefits on general benchmarks (average +2.53%), especially with substantial gains on MMStar (+7.26%), demonstrating the significance of the hierarchical thinking pattern and its effectiveness in enhancing MLLM capabilities. The project page is at https://vcu-bridge.github.io .
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。
