VIPL Team won the 1st place in two tracks of Visual Object Tracking (VOT) Challenge in ECCV 2022

Time: Nov 02, 2022

The 10th Visual Object Tracking (VOT) Challenge claimed the results at ECCV2022 in Tel Aviv, Israel, 24th October. VIPL Team won the 1st place in two tracks.

The VOT competition is the most influential competition in the field of visual object tracking, held annually at the international conference ICCV or ECCV since 2013, and continues to promote the development of tracking community. The competition currently includes several different tracks, such as short-time tracking (bounding-box and segmentation annotation), real-time tracking, and long-time tracking. The goal of the tracking task is that given the object to be tracked in the first frame of the video, the algorithm finds the corresponding target in each subsequent frames. The VIPL Team (Master student Botao Ye, Prof. Hong Chang, Prof. Bingpeng Ma, Prof. Shiguang Shan, Prof. Xilin Chen) participated in the competition. By adding additional temporal information to their tracking model OSTrack published in ECCV 2022, the ability to cope with distractors is further improved while real-time speed being guaranteed. Finally, the team won the first place in the tracks of VOT-STb2022 (box GT) & VOT-RTb2022.

Award certificates