Dataset-UAV-123 Video Dataset

Dataset-UAV-123 is a dataset composed of videos captured by low-altitude drones, which is fundamentally different from the videos in mainstream tracking datasets such as OTB50 and VOT2014. A subset of this dataset is used for long-term aerial tracking. It contains a total of 123 video sequences and more than 110k frames. It is also the second largest object tracking dataset after ALOV 300++.

All sequences in the UAV-123 Dataset are fully annotated with upright bounding boxes and can be easily integrated with the visual tracker benchmark, which contains all bounding box and attribute annotations of the UAV dataset.

This dataset was released by Matthias Mueller, Neil Smith and Bernard Ghanem of King Abdullah University of Science and Technology in 2016. The related paper is "A Benchmark and Simulator for UAV Tracking".