About LaSOT

Large-scale Single Object Tracking (LaSOT) aims to provide a dedicated platform for training data-hungry deep trackers as well as assessing long-term tracking performance. LaSOT is featured in

  • Large-scale: 1,400 sequences with more 3.52 millions frames
  • High-quality: Manual annotation with careful inspection in each frame
  • Category balance: 70 categories with each containing twenty sequences
  • Long-term tracking: An average video length of 2512 frames (i.e., 83 seconds)
  • Comprehensive labeling: providing both visual and lingual annotation for each sequence

Please check out the benchmark details and download links at LaSOT Benchmark page, evaluation toolkit and sample results at Evaluation and Result


H. Fan*, L. Lin*, F. Yang*, P. Chu*, G. Deng, S. Yu, H. Bai, Y. Xu, C. Liao, and H. Ling. LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking, CVPR, 2019. [Paper]

Core Team


We appreciate questions and suggestions to Heng Fan at hengfan@temple.edu or Haibin Ling at at hbling@temple.edu.