Abstract
Long-term language-guided referring in fixed-view videos is challenging: the referent may be occluded or leave the scene for long intervals and later re-enter, while framewise referring pipelines drift as re-identification (ReID) becomes unreliable. AR2-4FV leverages background stability for long-term referring. An offline Anchor Bank is distilled from static background structures; at inference, the text query is aligned with this bank to produce an Anchor Map that serves as persistent semantic memory when the referent is absent. An anchor-based re-entry prior accelerates re-capture upon return, and a lightweight ReID-Gating mechanism maintains identity continuity using displacement cues in the anchor frame. The system predicts per-frame bounding boxes without assuming the target is visible in the first frame or explicitly modeling appearance variations. AR2-4FV achieves +10.3% Re-Capture Rate (RCR) improvement and -24.2% Re-Capture Latency (RCL) reduction over the best baseline, and ablation studies further confirm the benefits of the Anchor Map, re-entry prior, and ReID-Gating.
Abstract (translated)
长期的基于语言指导的对象在固定视角视频中的识别是一项挑战，因为目标对象可能会被遮挡或长时间离开场景并在之后重新进入。帧级参考流水线在这种情况下容易出现偏差，因为重识别（ReID）变得不可靠。AR2-4FV利用背景稳定性来进行长期引用。 该系统通过从静态背景结构中离线提取一个锚点库（Anchor Bank），在推理过程中，文本查询与这个银行对齐以生成持续的语义记忆——锚图（Anchor Map）。当目标对象不在场时，此锚图充当持久性的语义存储。基于锚点的再进入先验加速了返回后的重新捕获，并且轻量级的身份门控机制利用锚帧中的位移线索来保持身份连续性。该系统预测每一帧的目标边界框，而不假设目标在第一帧中可见或显式地建模外观变化。 AR2-4FV与最佳基线相比，在再捕捉率（RCR）上提高了10.3%，并且减少了24.2%的再捕捉延迟（RCL）。消融研究进一步确认了锚图、再进入先验和身份门控机制的好处。
URL
https://arxiv.org/abs/2603.07758
