面向无人机跟踪的低成本高效时序学习
Toward Low-Cost yet Effective Temporal Learning for UAV Tracking
提出以 PPF（每 FLOP 精度）衡量时序学习组件的性价比，并用轻量的代表性外观 token 传播与合并实现时序建模，可直接迁移到任何单流跟踪器或需要低成本时序聚合的视频任务中。
原文依据 we introduce a new evaluation metric, i.e., precision per FLOP (PPF)
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论时序信息的利用一直是跟踪领域的一个开放课题。然而，现有跟踪器往往为时序学习引入越来越多的输入或参数，阻碍了其在资源受限的无人机（UAV）上的部署。更重要的是，这带来了一个模糊性问题：性能提升究竟来自时序学习本身，还是来自增加的输入与参数。在本研究中，我们主张从兼顾性能增益与计算成本的更均衡视角来设计时序学习组件。为实现这一目标，我们引入了一个新的评测指标，即每 FLOP 精度（PPF）。PPF 用于量化时序学习组件在单位 FLOP 下所取得的跟踪精度增益，从而能够在这些组件之间进行公平且考虑效率的比较，并推动它们走向更高效的设计。基于该指标，我们提出了一种低成本且有效的时序学习（LETL）方法，用于建模上下文关系。该方法在视频流中持续传播并合并具有代表性的外观 token，使跟踪器能够以相对较低的成本高效捕捉目标的变化模式。我们将 LETL 方法集成到现有的单流框架中，从而构建了一个简单而有效的跟踪器 LETrack，用于鲁棒的无人机跟踪。大量实验证明了 LETrack 的优越性，并表明所提出的 LETL 方法取得了更高的 PPF 分数。代码见 https://github.com/GXNU-ZhongLab/LETrack。
英文原文摘要 1634 字符
The utilization of temporal information has always been an open topic in the tracking community. However, existing trackers tend to employ more and more inputs or parameters for temporal learning, hindering their deployment in resource-constrained unmanned aerial vehicles (UAVs). More importantly, this raises ambiguity whether the performance gains come from the temporal learning itself, or come from the increased inputs and parameters. In this study, we advocate designing temporal learning components from a more balanced perspective that jointly considers performance gains and computational costs. To achieve this goal, we introduce a new evaluation metric, i.e., precision per FLOP (PPF). The PPF is introduced to quantify the tracking precision gains achieved by temporal learning components per unit of FLOP, thus enabling fair and efficiency-aware comparisons among these components and driving them toward more efficient designs. Based on this metric, we propose a low-cost yet effective temporal learning (LETL) approach to model contextual relationships. This approach continuously propagates and merges representative appearance tokens in video streams, allowing the tracker to efficiently capture the changing patterns of targets with relatively low costs. We integrate the LETL approach into existing one-stream frameworks, thereby building a simple yet effective tracker, namely LETrack, for robust UAV tracking. Extensive experimental demonstrate the superiority of our LETrack, and show that the proposed LETL approach achieves higher PPF scores. The Code is available at https://github.com/GXNU-ZhongLab/LETrack.