Confidence-Triggered Detection: An Approach to Build Real-time Tracking-by-detection System

arXiv:1902.00615v2 Announce Type: replace
Abstract: Real-time object tracking necessitates a delicate balance between speed and accuracy, a challenge exacerbated by the computational demands of deep learning methods. In this paper, we propose Confidence-Triggered Detection (CTD), an innovative approach that strategically bypasses object detection for frames closely resembling intermediate states, leveraging tracker confidence scores. CTD not only enhances tracking speed but also preserves accuracy, surpassing existing tracking algorithms. Through extensive evaluation across various tracker confidence thresholds, we identify an optimal trade-off …

abstract accuracy arxiv balance build challenge computational confidence cs.cv deep learning detection intermediate object paper real-time speed tracking type