During your internship, you will develop an innovative, data-driven, differentiable intrinsic camera-projection model alongside a dedicated methodology for its creation and application
You will implement this architecture as a differentiable neural network to facilitate end-to-end optimizable integration of camera calibration into downstream tasks
Additionally, you will place a strong emphasis on Deep-Learning frameworks such as 3D-Reconstruction, object detection, and V-SLAM
Finally, you will have the prospect of continuing this topic as a Master's thesis and publishing your developed methodology at a relevant conference.
Requisiti principali
Education: Master studies in a technical field like Computer Science or comparable
Experience and Knowledge: in Machine Learning, Linear Algebra and optionally in Computer Vision; first experience with Python (or equivalent programming language)
Personality and Working Practice: you excel at structuring your tasks independently and proactively achieving your goals with strong personal drive
Work Routine: presence in the office occasionally expected (at least once a month)
