VideoSearch-R1：透過軟性查詢精煉的迭代式影片檢索與推理

摘要

隨著视频语料库在规模和任务复杂性上持續增長，对于能够从大规模语料库中检索相关视频（视频间推理）并在检索内容中进行细粒度、查询条件驱动的任务（视频内推理，例如时间定位）的方法需求日益增加。然而，现有方法通常将检索视为预处理步骤，因此当初始检索失败时，缺乏改进搜索的机制，导致后续细粒度视频内推理的失败。此外，尽管近期的智能框架推动了视频理解的发展，但它们通常假设查询相关的视频已经给定，只专注于视频内推理任务。为了解决这些限制，我们提出VideoSearch-R1，一个通过与视频搜索引擎进行多轮交互实现迭代视频检索与推理的智能框架。具体而言，我们引入软查询精化（SQR），在连续潜在空间中精化搜索查询令牌，而非在离散文本空间中重写查询，从而实现更高效、更细粒度的调整。SQR及其推理过程通过群组相对策略优化（GRPO）进行训练，由从检索和下游任务中获得的任務级奖励信号引导。在此基础上，VideoSearch-R1在视频语料库时刻检索（VCMR）的三个数据集上达到最先进性能，能够迭代地从大规模语料库检索视频、精化搜索查询，并在检索内容中进行精确的查询条件时间定位。我们的分析表明，SQR有效精化了原始查询，所需生成的令牌数量远少于显式的文本级查询精化。代码和模型检查点已在mlvlab.github.io/VideoSearch-R1公开。

English

As video corpora continue to expand in both scale and task complexity, there is increasing demand for approaches that retrieve relevant videos from large-scale corpora (inter-video reasoning) and subsequently perform fine-grained, query-conditioned tasks (intra-video reasoning) within the retrieved content, such as temporal grounding. However, existing approaches typically treat retrieval as a preprocessing step, and consequently, when the initial retrieval fails, there is no mechanism to refine the search, leading to the failure of subsequent fine-grained intra-video reasoning. Moreover, while recent agentic frameworks have advanced video understanding, they typically assume that the query-relevant video is already given, focusing exclusively on intra-video reasoning tasks. To address these limitations, we propose VideoSearch-R1, an agentic framework for iterative video retrieval and reasoning through multi-turn interaction with a video search engine. Specifically, we introduce Soft Query Refinement (SQR) to refine search query tokens in a continuous latent space rather than rewriting queries in the discrete text space, enabling more efficient and fine-grained adjustments. SQR and its reasoning process are trained using Group Relative Policy Optimization (GRPO), guided by task-level reward signals derived from retrieval and downstream tasks. Building upon this, VideoSearch-R1 achieves state-of-the-art performance across three datasets on Video Corpus Moment Retrieval (VCMR), iteratively retrieving videos from large-scale corpora, refining search queries, and performing precise query-conditioned temporal grounding within the retrieved content. Our analyses show that SQR effectively refines the original query, requiring significantly fewer generated tokens than explicit text-level query refinement. Code and model checkpoints are publicly available at mlvlab.github.io/VideoSearch-R1.