LLOT

The LLOT dataset was created by Guilin University of Technology and collaborators to support target tracking research in low‑light environments. It contains 269 challenging video sequences, totaling over 132,000 frames, each meticulously annotated. The dataset aims to advance low‑light tracking technologies, addressing gaps in existing benchmarks. Applications span video surveillance, autonomous driving, medical imaging, and other domains requiring reliable tracking under dim conditions.