Abstract
Bounding box estimation by overlap maximization has improved the state of the art of visual tracking significantly, yet the improvement in robustness and accuracy is restricted by the limited reference information, i.e., the initial target. In this paper, we present DCOM, a novel bounding box estimation method for visual tracking, based on distribution calibration and overlap maximization. We assume every dimension in the modulation vector follows a Gaussian distribution, so that the mean and the variance can borrow from those of similar targets in large-scale training datasets. As such, sufficient and reliable reference information can be obtained from the calibrated distribution, leading to a more robust and accurate target estimation. Additionally, an updating strategy for the modulation vector is proposed to adapt the variation of the target object. Our method can be built on top of off-the-shelf networks without finetuning and extra parameters. It yields state-of-the-art performance on three popular benchmarks, including GOT-10k, LaSOT, and NfS while running at around 40 FPS, confirming its effectiveness and efficiency.
Highlights
Generic visual tracking is a long-standing topic in the field of computer vision and has attracted increasing attention in recent decades
We propose a novel bounding box estimation method for visual tracking, termed as DCOM, which is based on distribution calibration and overlap maximization
DiMP-DCOM runs at a slower speed compared with the bounding box regression (BBR) methods SiamBAN and SiamCAR, because they lack the process of online update
Summary
Generic visual tracking is a long-standing topic in the field of computer vision and has attracted increasing attention in recent decades. The single-object tracking task can be divided into two sub-tasks, i.e., localization and bounding box estimation, which aim at localizing the target roughly and predicting the precise bounding box, respectively. In order to build an accurate tracker, the bounding box estimation branch is of great importance, since it is responsible for generating the final bounding box directly. The previous works on bounding box estimation can be roughly grouped into three categories:. (1) multi-scale searching methods, (2) direct bounding box regression, and (3) bounding box estimation by overlap maximization. Conventional methods [12,13,14]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
