推动视觉语言模型面向跨域增量目标检测
Boosting Vision-Language Models Towards Cross-Domain Incremental Object Detection
把「跨域 + 新类别」同时纳入增量目标检测的评测协议，并用按分布分组的动态子空间 + 渐进式 adapter 合并来平衡稳定性与可塑性，可直接迁移到开放世界检测器或 VLM 检测器的持续适配场景。
原文依据 we propose Cross-Domain Incremental Object Detection (CDIOD), a new benchmark
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论增量目标检测（IOD）旨在使检测器具备应对动态环境和新兴目标类别的能力，而视觉语言模型的兴起大幅推进了这一目标。然而，现有研究往往假设增量任务来自单一通用域，从而过度简化了真实场景。为更好地在 IOD 下考察视觉语言模型，有必要探索同时涵盖新类别与新域的更泛化场景。为此，我们提出跨域增量目标检测（CDIOD），这是一个评估模型在不同域间持续适应多样目标检测任务能力的新基准。CDIOD 揭示出，在显著域偏移下，现有方法难以在自适应性与稳定性之间取得平衡。为应对这一挑战，我们提出动态分组子空间（DGS），这是一个新框架：它按分布对任务进行动态分组，以促进知识共享并防止任务冲突；渐进式地合并适配器以构建共享子空间并控制参数增长；并实现动态训练流程以维持恰当的稳定性—自适应性平衡。DGS 使视觉语言模型能够有效处理具有各种分布偏移的任务流。在三个基准上的大量实验表明，DGS 取得了 SOTA 性能，凸显了其在多样增量学习场景中的鲁棒性。代码已发布于 https://github.com/Never-wx/dgs。
英文原文摘要 1517 字符
Incremental Object Detection (IOD) aims to equip detectors with the ability to handle dynamic environments and emerging object categories, and the rise of vision-language models has substantially advanced this goal. However, existing studies often oversimplify real-world scenarios by assuming the incremental tasks come from a single general domain. To better investigate vision-language models under IOD, it is necessary to explore more generalized scenarios that encompass both novel categories and domains. To this end, we propose Cross-Domain Incremental Object Detection (CDIOD), a new benchmark that assesses the ability to continuously adapt to diverse object detection tasks across domains. CDIOD reveals that existing methods struggle to balance between adaptivity and stability under substantial domain shifts. To tackle this challenge, we propose Dynamic Group Subspace (DGS), a novel framework that dynamically groups tasks by distribution to promote knowledge sharing and prevent task collisions; progressively consolidates adapters to build shared subspaces and control parameter growth; and implements a dynamic training pipeline to maintain a proper stability-adaptivity balance. DGS enables vision-language models to effectively handle task streams of various distribution shifts. Extensive experiments across three benchmarks demonstrate that DGS achieves SOTA performance, highlighting its robustness in diverse incremental learning scenarios. Code is available at https://github.com/Never-wx/dgs.