Example-1. The swimmers in this photo are annotated.
Example-2. The swimmers in this photo are annotated.

The Swimmer Dataset was created for researchers aiming to train deep-learning networks capable of detecting swimmers or humans in open waters, with applications to sea emergency situations. To further simplify that process, the swimmers in the photos are annotated. Such a dataset was absent from the field and it is now publicly available and freely distributed, so any peer working on water search and rescue operations can benefit of its existence.

The dataset contains:

  • Photos of swimmers that can help train deep-learning networks
  • Text documents which include information about the annotations surrounding the swimmers

Click here to download dataset

If you use this dataset please cite:

  1. Eleftherios Lygouras, Nicholas Santavas, Anastasios Taitzoglou, Konstantinos Tarchanidis, Athanasios Mitropoulos and Antonios Gasteratos, “Unsupervised Human Detection with an Embedded Vision System on a Fully Autonomous UAV for Search and Rescue Operations”, Sensors, (MDPI), Vol 19, No 16, pp 3542, 2019, doi: 10.1109/LRA.2019.2897151.
  2. Eleftherios Lygouras, Antonios Gasteratos, Konstantinos Tarchanidis and Athanasios Mitropoulos, “ROLFER: A fully autonomous aerial rescue support system”, Microprocessors and Microsystems, (Elsevier), Vol. 61, pp. 32-42, 2018, doi: 10.1016/j.micpro.2018.05.014.3833.