Avalon - Aerial and vision-based assistance system for real time object detection in search and rescue missions

In this BMWi-funded project, multiple consortium partners work on developing an autonomous UAV to assist in search and rescue missions. Unmanned Aerial Vehicles (UAVs) equipped with cameras have emerged into an important asset in a wide range of fields, such as agriculture, delivery, surveillance, and search and rescue (SAR) missions. In particular, UAVs are capable of assisting in SAR missions due to their fast and versatile applicability while providing an overview over the scene. Especially in maritime scenarios, where wide areas need to be quickly overseen and searched, the efficient use of autonomous UAVs is crucial. Among the most challenging issues in this application scenario is the detection, localization, and tracking of people in open water. The small size of people relative to large search radii and the variability in viewing angles and altitudes require robust vision-based systems.

We develop computer vision systems for on- and off-board processing to automatically detect objects of interest. Please see the following works for further information:

[1] Benjamin Kiefer*, David Ott* and Andreas Zell. "Leveraging Synthetic Data in Object Detection on Unmanned Aerial Vehicles" , 2022 26th International Conference on Pattern Recognition (ICPR). (*equal contribution)
[2] Leon Amadeus Varga, Andreas Zell. "Tackling the Background Bias in Sparse Object Detection via Cropped Windows". In 2021  ICCV Workshop Vision Meets Drones, virtual, October 2021.
[3] L. A. Varga*, B. Kiefer*, M. Messmer*, and A. Zell, “SeaDronesSee: A Maritime Benchmark for Detecting Humans in Open Water,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), January 2022, pp. 2260–2270. (*equal contribution)
[4] Martin Meßmer*, Benjamin Kiefer* and Andreas Zell. "Gaining Scale Invariance in UAV Bird's Eye View Object Detection by Adaptive Resizing", 2022 26th International Conference on Pattern Recognition (ICPR). (*equal contribution)
[5] B. Kiefer*, M. Messmer* and A. Zell, "Diminishing Domain Bias by Leveraging Domain Labels in Object Detection on UAVs," 2021 20th International Conference on Advanced Robotics (ICAR), 2021, pp. 523-530, doi: 10.1109/ICAR53236.2021.9659357. (*equal contribution)

Datasets:

 

This project is funded by the German Ministry for Economic Affairs and Energy, Project Avalon, FKZ: 03SX481B.