Visual Tracking Using Drones

**Visual Tracking using Drones & Event cameras**:

- Réf
- **ABG-117654**
- Stage master 2 / Ingénieur- Durée 4 mois- Salaire net mensuel 4.05 (l'heure)- 07/11/2023- IMT Atlantique- Lieu de travail- Plouzané Bretagne France- Champs scientifiques- Informatique
- Mots clés- computer vision, robotics, deep learning, tracking, event sensors**Établissement recruteur**:
RAMBO team, Lab-STICC, IMT Atlantique, Brest (France)

**Description**:
Context:
This project will be conducted in the context of LEASARD project [1] which aims to increase the navigation autonomy of drones as Unmanned Aerial Vehicles (AUV) in search and rescue scenarios. Towards this goal, LEASARD project is dedicated to the enhancement of sensing and processing capabilities through the integration of event cameras and the utilization of appropriate deep neural networks to process data from these cameras.

Description and objectives:
Our primary focus in this project is to perform visual tracking using drones equipped with event cameras. While object tracking in standard videos has been extensively studied [2], the challenges are notably different in drone (aerial) videos. This is due to the fact that target objects are often smaller, and their appearances can change significantly due to variations in scale, orientation, and viewing angles [3].

To address these challenges, we plan to leverage newly emerging bio-inspired sensors known as Dynamic Vision Sensors (DVS) or event cameras [4]. In contrast to conventional RGB cameras, DVS sensors capture motion information with mínimal latency (in the order of microseconds), are almost free from motion blur and require less energy and computational resources. Additionally, DVS sensors outperform conventional cameras in terms of dynamic range (140 dB versus 60 dB), allowing them to operate effectively even in adverse lighting or weather conditions. Thanks to these unique properties, there has been a growing interest in harnessing these new sensors for autonomous drone navigation.

References:
[2] Marvasti-Zadeh, S. M., Cheng, L., Ghanei-Yakhdan, H., & Kasaei, S. (2021). Deep learning for visual tracking: A comprehensive survey. IEEE Transactions on Intelligent Transportation Systems, 23(5), 3943-3968.

[3] Hamdi, A., Salim, F., & Kim, D. Y. (2020, July). Drotrack: High speed drone-based object tracking under uncertainty. In 2020 IEEE international conference on fuzzy systems (FUZZ-IEEE) (pp. 1-8). IEEE.

[4] Gallego, G., Delbrück, T., Orchard, G., Bartolozzi, C., Taba, B., Censi, A.,... & Scaramuzza, D. (2020). Event-based vision: A survey. IEEE transactions on pattern analysis and machine intelligence, 44(1), 154-180.

[5] Wang, X., Li, J., Zhu, L., Zhang, Z., Chen, Z., Li, X.,... & Wu, F. (2021). Visevent: Reliable object tracking via collaboration of frame and event flows. IEEE Transactions on Cybernetics 2021.

[6] Zhu, P., Wen, L., Du, D., Bian, X., Fan, H., Hu, Q., & Ling, H. (2021). Detection and tracking meet drones challenge. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(11), 7380-7399.

**Profil**:

- A strong level of Python programming is required.
- An interest in deep learning frameworks (Pytorch) is also required.
- Good oral and written communication skills in English.

**Prise de fonction**:

- Dès que possible