释放层次推理：一种由LLM驱动的免训练指代视频目标分割框架
Unleashing Hierarchical Reasoning: An LLM-Driven Framework for Training-Free Referring Video Object Segmentation
AI总结：
提出免训练框架PARSE-VOS，利用LLM进行层次化粗到细推理，通过解析查询、时空定位和两阶段识别，在三个基准上达到最先进性能。
摘要：
指代视频目标分割（RVOS）旨在根据语言描述在视频中分割出感兴趣的目标。其突出挑战在于将静态文本与动态视觉内容对齐，特别是当具有相似外观的物体表现出不一致的运动和姿态时。然而，当前方法通常依赖整体性的视觉-语言融合，难以处理复杂、组合性的描述。在本文中，我们提出PARSE-VOS，一种新颖的、由大型语言模型（LLMs）驱动的免训练框架，用于跨文本和视频领域的层次化、从粗到细的推理。我们的方法首先将自然语言查询解析为结构化的语义命令。接下来，我们引入一个时空定位模块，在解析语义的引导下，为所有潜在目标对象生成所有候选轨迹。最后，一个层次化识别模块通过两阶段推理过程选择正确的目标：它首先使用LLM进行粗粒度的运动推理以缩小候选范围；如果仍存在歧义，则条件触发细粒度的姿态验证阶段以消除歧义。最终输出是目标对象的准确分割掩码。PARSE-VOS在三个主要基准上取得了最先进的性能：Ref-YouTube-VOS、Ref-DAVIS17和MeViS。
英文摘要：
Referring Video Object Segmentation (RVOS) aims to segment an object of interest throughout a video based on a language description. The prominent challenge lies in aligning static text with dynamic visual content, particularly when objects exhibiting similar appearances with inconsistent motion and poses. However, current methods often rely on a holistic visual-language fusion that struggles with complex, compositional descriptions. In this paper, we propose \textbf{PARSE-VOS}, a novel, training-free framework powered by Large Language Models (LLMs), for a hierarchical, coarse-to-fine reasoning across text and video domains. Our approach begins by parsing the natural language query into structured semantic commands. Next, we introduce a spatio-temporal grounding module that generates all candidate trajectories for all potential target objects, guided by the parsed semantics. Finally, a hierarchical identification module select the correct target through a two-stage reasoning process: it first performs coarse-grained motion reasoning with an LLM to narrow down candidates; if ambiguity remains, a fine-grained pose verification stage is conditionally triggered to disambiguate. The final output is an accurate segmentation mask for the target object. \textbf{PARSE-VOS} achieved state-of-the-art performance on three major benchmarks: Ref-YouTube-VOS, Ref-DAVIS17, and MeViS.
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。
