The VOT 2015 and VOT-TIR 2015 Challenges Submission Report

In this report, we present the details of our submission and the evaluation results on both the visual object tracking (VOT) 2015 and the thermal infrared visual object tracking (VOT-TIR) 2015 challenges. The goal of these challenges is to compare short-term model-free single-object trackers, and serve as the de factor state-of-the-art evaluation platform for visual object tracking. In particular, the VOT challenge focuses on natural RGB video sequences with rotated rectangle ground truth boxes, while the VOT-TIR challenge consists of thermal infrared video sequences with axis-aligned ground truth boxes, see examples in Figures 1 and 2. For more details of the two challenges, we refer the reader to the official challenge reports [3, 1].

[1]  C. Lawrence Zitnick,et al.  Edge Boxes: Locating Object Proposals from Edges , 2014, ECCV.

[2]  Michael Felsberg,et al.  The Thermal Infrared Visual Object Tracking VOT-TIR2015 Challenge Results , 2015, ICCV Workshops.

[3]  Cordelia Schmid,et al.  Online Object Tracking with Proposal Selection , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[4]  Michael Felsberg,et al.  The Visual Object Tracking VOT2015 Challenge Results , 2015, 2015 IEEE International Conference on Computer Vision Workshop (ICCVW).