Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation

TL;DR

Alpha-Refine introduces a novel refinement module with pixel-wise correlation and spatial-aware non-local layers, significantly enhancing bounding box accuracy in visual tracking when integrated with state-of-the-art trackers.

Contribution

The paper presents Alpha-Refine, a flexible refinement module that improves tracking accuracy by combining pixel-wise correlation and non-local features, applicable to multiple base trackers.

Findings

Significant performance improvements on TrackingNet, LaSOT, and VOT2018 benchmarks.

Effective fusion of multiple outputs (bounding box, corners, mask) for better tracking.

Compatibility with various state-of-the-art trackers enhances their precision.

Abstract

In recent years, the multiple-stage strategy has become a popular trend for visual tracking. This strategy first utilizes a base tracker to coarsely locate the target and then exploits a refinement module to obtain more accurate results. However, existing refinement modules suffer from the limited transferability and precision. In this work, we propose a novel, flexible and accurate refinement module called Alpha-Refine, which exploits a precise pixel-wise correlation layer together with a spatial-aware non-local layer to fuse features and can predict three complementary outputs: bounding box, corners and mask. To wisely choose the most adequate output, we also design a light-weight branch selector module. We apply the proposed Alpha-Refine module to five famous and state-of-the-art base trackers: DiMP, ATOM, SiamRPN++, RTMDNet and ECO. The comprehensive experiments on TrackingNet, LaSOT and VOT2018 benchmarks demonstrate that our approach significantly improves the tracking performance in comparison with other existing refinement methods. The source codes will be available at https://github.com/MasterBin-IIAU/AlphaRefine.