Efficient Language-to-Vision Feature Injection for Referring Single-Object Tracking

2026-08-29

AI Summary

本文提出LVTrack，通过模式条件门控特征注入器和预训练模型来解决语言到视觉特征的有效注入问题，减少训练成本并提高单目标跟踪性能。

Abstract

Referring single-object tracking enables language-grounded target initialization and subsequent tracking by jointly leveraging semantic cues and visual templates. The core difficulty is to use language differently across stages: it is indispensable for grounding but can induce semantic drift during tracking when overemphasized. Meanwhile, current methods often require costly vision-language alignment training. We present LVTrack, a pure transformer framework that introduces a mode-conditioned Gated Feature Injector to adaptively regulate textual guidance and alleviate semantic drift. Together with targeted adaptations, it directly harnesses a frozen vision-language pretrained model, greatly reducing training cost and preserving strong language understanding. To further improve temporal localization, LVTrack integrates hybrid relative-absolute positional encodings with a lightweight memory mechanism and optimizes autoregressive box prediction using a Gaussian-smoothed KL loss. Extensive experiments on standard benchmarks demonstrate that LVTrack achieves strong performance.

Problem

• Referring single-object tracking
• language-grounded target initialization
• semantic drift
• vision-language alignment training

Innovation

• Gated Feature Injector
• vision-language pretrained model
• hybrid positional encodings
• lightweight memory mechanism
• Gaussian-smoothed KL loss