用轻量时序注意力统一目标追踪与轨迹预测，实时稳定且内存固定。
SOTFormer: A Minimal Transformer for Unified Object Tracking and Trajectory Prediction
- 单层时序注意力+真值引导记忆，实现身份稳定追踪。
- 在Mini-LaSOT上达76.3 AUC、53.7 FPS，内存仅4.3 GB。
- 适合需要实时性与抗遮挡的视觉感知系统部署。
准确的单目标追踪与短期运动预测在遮挡、尺度变化和时间漂移下仍具挑战，破坏了实时感知所需的时序一致性。我们提出SOTFormer，一种最小化常量内存的时序变换器，将目标检测、追踪与短时轨迹预测统一于单一端到端框架中。不同于具有循环或堆叠时序编码器的先前模型，SOTFormer通过真值引导的记忆与烧入锚点损失，显式稳定初始化，实现稳定的身份传播。单个轻量级时序注意力层在帧间精炼特征表示，支持实时推理且GPU内存固定。在Mini-LaSOT（20%）基准上，SOTFormer取得76.3 AUC与53.7 FPS（AMP，4.3 GB VRAM），优于TrackFormer与MOTRv2等变换器基线，在快速运动、尺度变化与遮挡条件下表现更优。
原文摘要 · Abstract (English)
Accurate single-object tracking and short-term motion forecasting remain challenging under occlusion, scale variation, and temporal drift, which disrupt the temporal coherence required for real-time perception. We introduce \textbf{SOTFormer}, a minimal constant-memory temporal transformer that unifies object detection, tracking, and short-horizon trajectory prediction within a single end-to-end framework. Unlike prior models with recurrent or stacked temporal encoders, SOTFormer achieves stable identity propagation through a ground-truth-primed memory and a burn-in anchor loss that explicitly stabilizes initialization. A single lightweight temporal-attention layer refines embeddings across frames, enabling real-time inference with fixed GPU memory. On the Mini-LaSOT (20%) benchmark, SOTFormer attains 76.3 AUC and 53.7 FPS (AMP, 4.3 GB VRAM), outperforming transformer baselines such as TrackFormer and MOTRv2 under fast motion, scale change, and occlusion.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。