🛠 Tech Stack

💼 About This Role

You'll join the Perception team to develop a real-time video-based detector and integrate it with a tracking algorithm for autonomous UAVs. Your work will directly enable drones to "see" and "follow" targets in flight tests.

🎯 What You'll Do

• Design and train video-based object detection models (e.g., YOLO, lightweight Transformers).
• Integrate detection into tracking algorithms (Sort/DeepSort, CSRT) for high-dynamics lock.
• Optimize models with TensorRT for real-time inference on embedded hardware.
• Participate in field testing to validate algorithms in diverse environments.

📋 Requirements

• Currently pursuing or completed a Master's degree in CS, Robotics, EE, or related field.
• Strong understanding of CNN architectures and object detection frameworks.
• Proficiency in Python (PyTorch/TensorFlow) and comfortable with C++.
• Experience with Linux environment and embedded systems.

✨ Nice to Have

• Experience with NVIDIA Jetson platforms and hardware-accelerated inference.
• FPV pilot experience or hobbyist interest in UAVs.
• Experience with synthetic data generation (NVIDIA Isaac Sim, Gazebo).

🎁 Benefits & Perks

• 💻 Hands-on experience with cutting-edge defense technology
• 🚀 Work on live flight tests
• 🧑‍🏫 Mentorship from senior engineers