期刊
IEEE ROBOTICS AND AUTOMATION LETTERS
卷 8, 期 2, 页码 1133-1140出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/LRA.2023.3236583
关键词
Autonomous aerial vehicles; Target tracking; Three-dimensional displays; Transformers; Correlation; Training; Trajectory; Unmanned aerial vehicle; visual object tracking; voxel-based trajectory-aware pre-training; self-supervised learn- ing; hierarchical self-attention Transformer
类别
This research presents TRTrack, a comprehensive framework that fully utilizes stereoscopic representation for UAV tracking. Through trajectory-aware reconstruction training (TRT) and spatial correlation refinement (SCR), the framework improves the performance of UAV tracking.
Siamese network-based object tracking has remarkably promoted the automatic capability for highly-maneuvered unmanned aerial vehicles (UAVs). However, the leading-edge tracking framework often depends on template matching, making it trapped when facing multiple views of object in consecutive frames. Moreover, the general image-level pretrained backbone can overfit to holistic representations, causing the misalignment to learn object-level properties in UAV tracking. To tackle these issues, this work presents TRTrack, a comprehensive framework to fully exploit the stereoscopic representation for UAV tracking. Specifically, a novel pre-training paradigm method is proposed. Through trajectory-aware reconstruction training (TRT), the capability of the backbone to extract stereoscopic structure feature is strengthened without any parameter increment. Accordingly, an innovative hierarchical self-attention Transformer is proposed to capture the local detail information and global structure knowledge. For optimizing the correlation map, we proposed a novel spatial correlation refinement (SCR) module, which promotes the capability of modeling the long-range spatial dependencies. Comprehensive experiments on three UAV challenging benchmarks demonstrate that the proposed TRTrack achieves superior UAV tracking performance in both precision and efficiency. Quantitative tests in real-world settings fully prove the effectiveness of our work.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据