Layer-Guided UAV Tracking: Enhancing Efficiency and Occlusion Robustness

2026-02-14

AI Summary

This work addresses the challenge of simultaneously achieving high accuracy and efficiency in visual object tracking for unmanned aerial vehicles (UAVs), particularly under adverse conditions such as occlusion where robustness is often compromised. To this end, the authors propose LGTrack, a lightweight tracking framework that introduces a novel Light-weight Global Group Coordinate Attention (GGCA) module to effectively capture long-range dependencies and global contextual information. Furthermore, a Similarity-Guided Layer Adaptation (SGLA) mechanism is devised to enable efficient, adaptive feature layer selection, serving as a more effective alternative to conventional knowledge distillation. The resulting architecture achieves an optimal balance between precision and computational efficiency, delivering state-of-the-art performance with 82.8% precision at a real-time inference speed of 258.7 FPS across three benchmarks, including UAVDT, significantly outperforming existing approaches.

Technology Category

• Computer Vision: Motion & Tracking
• Search and Optimization: Learning to Search
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Visual object tracking (VOT) plays a pivotal role in unmanned aerial vehicle (UAV) applications. Addressing the trade-off between accuracy and efficiency, especially under challenging conditions like unpredictable occlusion, remains a significant challenge. This paper introduces LGTrack, a unified UAV tracking framework that integrates dynamic layer selection, efficient feature enhancement, and robust representation learning for occlusions. By employing a novel lightweight Global-Grouped Coordinate Attention (GGCA) module, LGTrack captures long-range dependencies and global contexts, enhancing feature discriminability with minimal computational overhead. Additionally, a lightweight Similarity-Guided Layer Adaptation (SGLA) module replaces knowledge distillation, achieving an optimal balance between tracking precision and inference efficiency. Experiments on three datasets demonstrate LGTrack's state-of-the-art real-time speed (258.7 FPS on UAVDT) while maintaining competitive tracking accuracy (82.8\% precision). Code is available at https://github.com/XiaoMoc/LGTrack

Problem

• UAV tracking
• occlusion robustness
• efficiency-accuracy trade-off
• visual object tracking

Innovation

• Layer-Guided Tracking
• Global-Grouped Coordinate Attention (GGCA)
• Similarity-Guided Layer Adaptation (SGLA)
• Occlusion Robustness
• Real-Time UAV Tracking