Dynamic Semantic-Aware Correlation Modeling for UAV Tracking

2025-10-24

AI Summary

Existing UAV tracking methods prioritize speed but neglect semantic awareness, leading to inaccurate localization feature extraction between template and search regions—especially under challenges such as camera motion, rapid target movement, and low-resolution input. To address this, we propose a Dynamic Semantic-Aware Correlation Modeling framework. First, we design a Transformer-driven Correlation Map Enhancement module to strengthen the search region’s response to semantically critical template features. Second, we introduce a Dynamic Semantic Correlation Generator that jointly optimizes semantic understanding and correlation modeling. Third, we employ structured pruning to construct multiple model variants, enabling flexible trade-offs between accuracy and efficiency. Our approach achieves significant improvements in both accuracy and robustness across multiple UAV tracking benchmarks. The source code is publicly available.

Technology Category

• Computer Vision: Motion & Tracking
• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

UAV tracking can be widely applied in scenarios such as disaster rescue, environmental monitoring, and logistics transportation. However, existing UAV tracking methods predominantly emphasize speed and lack exploration in semantic awareness, which hinders the search region from extracting accurate localization information from the template. The limitation results in suboptimal performance under typical UAV tracking challenges such as camera motion, fast motion, and low resolution, etc. To address this issue, we propose a dynamic semantic aware correlation modeling tracking framework. The core of our framework is a Dynamic Semantic Relevance Generator, which, in combination with the correlation map from the Transformer, explore semantic relevance. The approach enhances the search region's ability to extract important information from the template, improving accuracy and robustness under the aforementioned challenges. Additionally, to enhance the tracking speed, we design a pruning method for the proposed framework. Therefore, we present multiple model variants that achieve trade-offs between speed and accuracy, enabling flexible deployment according to the available computational resources. Experimental results validate the effectiveness of our method, achieving competitive performance on multiple UAV tracking datasets. The code is available at https://github.com/zxyyxzz/DSATrack.

Problem

• Improves UAV tracking accuracy under camera motion
• Enhances semantic relevance modeling for better localization
• Balances tracking speed and accuracy through pruning

Innovation

• Dynamic Semantic Relevance Generator for correlation modeling
• Transformer-based semantic relevance exploration for tracking
• Pruning method enables speed-accuracy trade-off optimization