TGTrack：面向统一单目标跟踪的时序生成式学习
TGTrack: Temporal Generative Learning for Unified Single Object Tracking
把「预测未来帧特征」作为自监督辅助任务来训练跟踪器，并用时间词元显式编码帧位置；这套生成式时序监督可直接迁移到任何需要建模帧间动态的视觉跟踪或视频表征学习框架中。
原文依据 a temporally generative supervision task to guide the model in learning temporal dynamics
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论现有的单目标跟踪器通常对时序建模的处理较为表面，仅传递有限的帧间信息，例如传播词元或模板更新，而缺乏内在的时序监督学习。为解决这一局限，我们提出 TGTrack，一种新的统一跟踪框架，它引入一个时序生成式监督任务来引导模型学习时序动态。TGTrack 的核心是一种时序生成式学习范式，配备基于 Transformer 的生成式解码器，该解码器由门控融合模块和自回归预测机制组成。这一联合设计使模型能够根据先前信息推断未来场景，从而提升其对视觉外观与时序动态的建模能力。此外，我们引入时间词元嵌入，以显式编码每一帧的时序位置。在涵盖五种模态的 11 个基准上的实验表明，TGTrack 在鲁棒统一跟踪方面达到最优性能。例如，TGTrack-B384 在 LaSOT 上取得 75.3% 的 AUC。代码见 https://github.com/wtg1/TGTrack。
英文原文摘要 1148 字符
Existing single object trackers typically treat temporal modeling superficially by passing limited inter-frame information, such as propagated tokens or template updates, without intrinsic temporal supervision learning. To address this limitation, we propose TGTrack, a new unified tracking framework that incorporates a temporally generative supervision task to guide the model in learning temporal dynamics. The core of TGTrack is a temporally generative learning paradigm equipped with a transformer-based generative decoder, which consists of a gated fusion module and an autoregressive prediction mechanism. This joint design enables the model to infer future scenarios from preceding information, thereby improving its ability to model both visual appearance and temporal dynamics. Furthermore, we introduce a time token embedding to explicitly encode the temporal position of each frame. Experiments on 11 benchmarks spanning five modalities show that TGTrack achieves state-of-the-art performance in robust unified tracking. For instance, TGTrack-B384 achieves an AUC of 75.3% on LaSOT. Code is available at https://github.com/wtg1/TGTrack.