VOTS2023 Benchmark

Tracking has matured to a point where the constraints enforced in past VOT challenges can be relaxed and general object tracking should be considered in a broader context. The new Visual Object Tracking and Segmentation challenge VOTS2023 thus no longer distincts between single- and multi-target tracking nor between short- and long-term tracking. We propose a single challenge that requires tracking one or more targets simultaneously by segmentation over long or short sequences, while the targets may disappear during tracking and reappear later in the video.

The VOTS2023 Benchmark is continually open for post-challenge evaluation, see the leaderboard for up-to-date results. Please follow instructions to submit your results. The paper with the results of the challenge is available here, the presentation with the results is available here.

What’s new

  • A single challenge unifies single/multi-target short/long-term segmentation tracking tasks.
  • A new multi-target segmentation tracking dataset.
  • A new performance evaluation protocol and measures.
  • All trackers will be evaluated on a sequestered ground truth.
  • Participate by running your single-target or multiple-target tracker (same challenge)

Sponsors

The VOTS2023 workshop is sposored by the Faculty of Computer and Information Science, University of Ljubljana, Eyedea, and University of Birmingham.