Explicit Context Reasoning with Supervision for Visual Tracking

TL;DR

This paper introduces RSTrack, a novel visual tracking method that explicitly models and supervises context reasoning to improve temporal consistency and target representation accuracy, achieving state-of-the-art results.

Contribution

Proposes RSTrack, which explicitly models and supervises context reasoning with three mechanisms, enhancing temporal consistency and robustness in visual tracking.

Findings

Achieves state-of-the-art performance on multiple benchmarks.

Maintains real-time tracking speeds.

Effectively suppresses drift in context reasoning.

Abstract

Contextual reasoning with constraints is crucial for enhancing temporal consistency in cross-frame modeling for visual tracking. However, mainstream tracking algorithms typically associate context by merely stacking historical information without explicitly supervising the association process, making it difficult to effectively model the target's evolving dynamics. To alleviate this problem, we propose RSTrack, which explicitly models and supervises context reasoning via three core mechanisms. \textit{1) Context Reasoning Mechanism}: Constructs a target state reasoning pipeline, converting unconstrained contextual associations into a temporal reasoning process that predicts the current representation based on historical target states, thereby enhancing temporal consistency. \textit{2) Forward Supervision Strategy}: Utilizes true target features as anchors to constrain the reasoning pipeline, guiding the predicted output toward the true target distribution and suppressing drift in the context reasoning process. \textit{3) Efficient State Modeling}: Employs a compression-reconstruction mechanism to extract the core features of the target, removing redundant information across frames and preventing ineffective contextual associations. These three mechanisms collaborate to effectively alleviate the issue of contextual association divergence in traditional temporal modeling. Experimental results show that RSTrack achieves state-of-the-art performance on multiple benchmark datasets while maintaining real-time running speeds. Our code is available at https://github.com/GXNU-ZhongLab/RSTrack.