Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Subjects: | |
| Online Access: | https://doi.org/10.5281/zenodo.17182190 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <h2><strong>Associated Paper: <em>Detector-Augmented SAMURAI for Long-Duration Drone Tracking</em></strong></h2> <h3><strong>Published in the Proceedings of the 2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops!</strong></h3> <p>The <em>Long-Duration Drone Tracking Dataset</em> is part of the paper <strong>"Detector-Augmented SAMURAI for Long-Duration Drone Tracking" </strong>by Tamara R. Lenhard, Andreas Weinmann, Hichem Snoussi, Tobias Koch. This work was presented in the 2026 IEEE/CVF WACV Workshop on <em>Real-World Surveillance: Applications and Challenges (RWS) </em>and is published in the Proceedings of the 2026 IEEE/CVF WACV Workshops.</p> <p><strong>Final Version: <a href="https://openaccess.thecvf.com/content/WACV2026W/RWS/html/Lenhard_Detector-Augmented_SAMURAI_for_Long-Duration_Drone_Tracking_WACVW_2026_paper.html">here</a></strong></p> <p>(ArXiv Preprint:<strong> </strong><a href="https://arxiv.org/abs/2601.04798">here</a>)</p> <p> </p> <h2><strong>Dataset Details</strong></h2> <p>The Long-Duration Drone Tracking Dataset consists of two subsets, <strong>R1</strong> and <strong>R2</strong>, each containing two long-duration sequences. Every sequence is recorded at a resolution of <strong>2040 × 1086 pixels</strong> using a ground-mounted <strong>Basler acA200-165c</strong> camera system equipped with two different lenses (R1: 8 mm lens, R2: 25 mm lens). The recordings were captured in an urban environment characterized by medium-density vegetation and medium-height buildings.</p> <p><strong>Annotation Format:</strong> Annotations (bounding boxes) are provided via text files according to the <strong>YOLO standard format</strong></p> <pre><code><object-class> <x> <y> <width> <height></code></pre> <p>Here, <code><x></code> and <code><y></code> represent the normalized coordinates of the bounding box center, while <code><width></code> and <code><height></code> denote the normalized bounding box wisth and height. In SynDroneVision, <code><object-class></code> is always 0, indicating the drone class.</p> <h2><strong>Download</strong></h2> <p>Image data and annotations are available via the following links:</p> <table style="border-collapse: collapse; width: 80%; height: 98px; border-width: 1px; margin-left: auto; margin-right: auto;"><colgroup><col style="width: 18.2749%;"><col style="width: 14.9123%;"><col style="width: 13.748%;"><col style="width: 53.0648%;"></colgroup> <tbody> <tr style="height: 19.6px;"> <td style="height: 19.6px;">Subset Name</td> <td style="height: 19.6px;">Sequence</td> <td style="height: 19.6px;">Size </td> <td style="height: 19.6px;">Download Link</td> </tr> <tr style="height: 19.6px;"> <td style="height: 19.6px;">R1</td> <td style="height: 19.6px;">POS3</td> <td style="height: 19.6px;">7.1 GB</td> <td style="height: 19.6px;"><a href="https://datastore.dlr-pi.de/s/E7aoCn988jK9Hoo">Download R1_POS3</a></td> </tr> <tr style="height: 19.6px;"> <td style="height: 19.6px;">R1</td> <td style="height: 19.6px;">POS7</td> <td style="height: 19.6px;">4.7 GB</td> <td style="height: 19.6px;"><a href="https://datastore.dlr-pi.de/s/5Xp7JHeZn4WwCp2" target="_blank" rel="noopener">Download R1_POS7</a></td> </tr> <tr style="height: 19.6px;"> <td style="height: 19.6px;">R2</td> <td style="height: 19.6px;">POS3</td> <td style="height: 19.6px;">1.8 GB</td> <td style="height: 19.6px;"><a href="https://datastore.dlr-pi.de/s/eYN8xJZY3rdmqm7" target="_blank" rel="noopener">Download R2_POS3</a></td> </tr> <tr style="height: 19.6px;"> <td style="height: 19.6px;">R2</td> <td style="height: 19.6px;">POS7</td> <td style="height: 19.6px;">5.4 GB</td> <td style="height: 19.6px;"><a href="https://datastore.dlr-pi.de/s/DQBtASX6oJ3t4RE" target="_blank" rel="noopener">Download R2_POS7</a></td> </tr> </tbody> </table> <p>Example images of each sequence are shown below!</p> <p> </p> <h2><strong>Citation</strong></h2> <p>If you find the Long-Duration Drone Tracking Dataset helpful in your research, we kindly ask that you cite the associated paper. Below is the citation in BibTeX format for your convenience:</p> <p><strong>BibTeX:</strong></p> <pre>@InProceedings{Lenhard_2026_WACV, <br> author = {Lenhard, Tamara R. and Weinmann, Andreas and Snoussi, Hichem and Koch, Tobias}, <br> title = {Detector-Augmented SAMURAI for Long-Duration Drone Tracking}, <br> booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops}, <br> month = {March}, <br> year = {2026}, <br> pages = {75-84} <br>} </pre>