建模跨视觉协同的统一大视觉模型
Modeling Cross-vision Synergy for Unified Large Vision Model
PolyV 的稀疏混合专家架构与动态模态路由器为视觉多模态统一模型提供了可复用的设计范式，其协同感知训练策略（模态特定预训练+知识蒸馏+对象/关系级对齐）可直接用于提升视觉模型在跨模态推理任务上的表现。
原文依据 PolyV adopts a sparse Mixture-of-Experts LVM coordinated by a dynamic modality router
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期大视觉模型（LVM）的进展已从模态特定设计转向联合处理图像、视频和3D数据的统一架构。然而，现有统一LVM主要追求功能集成，而忽视了跨视觉协同这一更深层目标：即跨视觉模态对互补先验进行推理的能力。为此，我们提出PolyV，一种在架构和训练层面均实现跨视觉协同的统一LVM。在架构上，PolyV采用稀疏混合专家LVM，由动态模态路由器协调，使每个专家专注于模态特定先验，同时实现跨模态的双向交互与相互精炼。在训练上，协同感知范式将模态特定预训练与通过知识蒸馏和对象/关系级对齐的由粗到细协同调优相结合。在涵盖图像、视频和3D理解的10个基准（包括需要空间或时序先验的协同聚焦数据集）上的大量实验表明，PolyV持续优于现有模型，相比其骨干网络平均提升超过10%。总体而言，PolyV为通感视觉推理建立了一个统一框架，推动迈向真正协同的LVM。
英文原文摘要 1334 字符
Recent advances in large vision models (LVMs) have shifted from modality-specific designs toward unified architectures that jointly process images, videos, and 3D data. However, existing unified LVMs primarily pursue functional integration, while overlooking the deeper goal of cross-vision synergy: the ability to reason over complementary priors across visual modalities. To address this, we present PolyV, a unified LVM that achieves cross-vision synergy at both the architectural and training levels. Architecturally, PolyV adopts a sparse Mixture-of-Experts LVM coordinated by a dynamic modality router, allowing each expert to specialize in modality-specific priors while enabling bidirectional interaction and mutual refinement across modalities. Training-wise, a synergy-aware paradigm combines modality-specific pretraining with coarse-to-fine synergy tuning via knowledge distillation and object-/relation-level alignment. Extensive experiments on 10 benchmarks spanning image, video, and 3D understanding, including synergy-focused datasets requiring spatial or temporal priors, demonstrate that PolyV consistently outperforms existing models, achieving over 10% average improvement over its backbone. Overall, PolyV establishes a unified framework for synesthetic visual reasoning, advancing toward truly synergistic LVMs.