Master Thesis - AI-based 2D Pixel-wise Tracking for Space Applications

We are looking for a motivated student (m/f/d) interested in computer vision and machine learningfor space applications on our On-Orbit-Servicing Simulator (OOS-SIM).Here, obtaining precise satellite pose estimations from camera data is crucial for close rangesatellite operations such as repair or refuel.The work focuses on developing a 2D instance segmentation tracking approach that is especiallysuited for the challenging space domain.

Your Tasks:

• Perform literature review on existing state-of-the-art pixel-wise tracking methods andsegmentation approaches (traditional, CNN, deep learning based, ..)
• Identify key challenges of working with space domain images
• Adapt (or develop a new) semantic image segmentation / pixel tracking approach forsatellite geometries in space
• Evaluate and compare the developed method on synthetic and Hardware-In-the-Loop (HIL)data on the OOS-SIM

Your Qualifications:

• Pursuing a Master’s degree in Robotics, Computer Science, Engineering, or a related field
• Strong proficiency in Python and deep learning frameworks (primarily PyTorch)
• Familiarity with image processing or synthetic data generation or uncertainty estimationmethods
• Experience with Linux environments and git is a significant advantage
• Fluency in English and/or German

Why DLR?:

• Work alongside leading scientists in a highly collaborative, international atmosphere
• Gain valuable hands-on experience with state-of-the-art vision models
• Deploy your work on robotic systems in our lab