Architecture and evaluation protocol for transformer-based visual object tracking in UAV applications
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917362047909888 |
|---|---|
| author | Borne, Augustin Notin, Pierre Hennequin, Christophe Changey, Sebastien Bazeille, Stephane Cudel, Christophe Quint, Franz |
| author_facet | Borne, Augustin Notin, Pierre Hennequin, Christophe Changey, Sebastien Bazeille, Stephane Cudel, Christophe Quint, Franz |
| contents | Object tracking from Unmanned Aerial Vehicles (UAVs) is challenged by platform dynamics, camera motion, and limited onboard resources. Existing visual trackers either lack robustness in complex scenarios or are too computationally demanding for real-time embedded use. We propose an Modular Asynchronous Tracking Architecture (MATA) that combines a transformer-based tracker with an Extended Kalman Filter, integrating ego-motion compensation from sparse optical flow and an object trajectory model. We further introduce a hardware-independent, embedded oriented evaluation protocol and a new metric called Normalized time to Failure (NT2F) to quantify how long a tracker can sustain a tracking sequence without external help. Experiments on UAV benchmarks, including an augmented UAV123 dataset with synthetic occlusions, show consistent improvements in Success and NT2F metrics across multiple tracking processing frequency. A ROS 2 implementation on a Nvidia Jetson AGX Orin confirms that the evaluation protocol more closely matches real-time performance on embedded systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_03904 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Architecture and evaluation protocol for transformer-based visual object tracking in UAV applications Borne, Augustin Notin, Pierre Hennequin, Christophe Changey, Sebastien Bazeille, Stephane Cudel, Christophe Quint, Franz Computer Vision and Pattern Recognition Object tracking from Unmanned Aerial Vehicles (UAVs) is challenged by platform dynamics, camera motion, and limited onboard resources. Existing visual trackers either lack robustness in complex scenarios or are too computationally demanding for real-time embedded use. We propose an Modular Asynchronous Tracking Architecture (MATA) that combines a transformer-based tracker with an Extended Kalman Filter, integrating ego-motion compensation from sparse optical flow and an object trajectory model. We further introduce a hardware-independent, embedded oriented evaluation protocol and a new metric called Normalized time to Failure (NT2F) to quantify how long a tracker can sustain a tracking sequence without external help. Experiments on UAV benchmarks, including an augmented UAV123 dataset with synthetic occlusions, show consistent improvements in Success and NT2F metrics across multiple tracking processing frequency. A ROS 2 implementation on a Nvidia Jetson AGX Orin confirms that the evaluation protocol more closely matches real-time performance on embedded systems. |
| title | Architecture and evaluation protocol for transformer-based visual object tracking in UAV applications |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.03904 |