LaSOT Large-Scale Single Object Tracking Dataset

LaSOT stands for Large-scale Single Object Tracking, a high-quality large-scale single object tracking benchmark. LaSOT consists of 1,400 sequences with a total of more than 3.5 million frames, and each frame is manually annotated with a bounding box, making LaSOT one of the largest densely annotated tracking benchmarks.

The videos in this dataset have an average length of over 2,500 frames, and each sequence includes complex in-the-wild challenges such as target objects disappearing and reappearing in view.