DLR.de/jobs
Course paper/final thesis
Enter the fascinating world of the German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt; DLR) and help shape the future through research and innovation! We offer an exciting and inspiring working environment driven by the expertise and curiosity of our 11,000 employees from 100 nations and our unique infrastructure. Together, we develop sustainable technologies and thus contribute to finding solutions to global challenges. Would you like to join us in addressing this major future challenge? Then this is your place!
For our Institute for the Protection of Terrestrial Infrastructures in Sankt Augustin we are looking for a
Student (f/m/x) Computer Science, Mathematics or similar
Real-time Light and Weather Adaption in Game Engines to support Object Detection using Synthetic Data
What to expect:
The lack of publicly available data, the cost intensity, and the high time expenditure associated with acquiring labeled data are leading to the increasing use of synthetic data when training deep neural networks (DNN) for diverse computer vison (CV) tasks (e.g., drone detection). Data generation methods range from fairly simple techniques, such as domain randomization, to game engine-based simulations in three-dimensional environments (e.g., via Unreal Engine). One important factor for the successful usage of synthetic data in the development of image-based detection techniques is the reconstruction of realistic weather and lighting conditions. While game engines provide plugins for setting the sun’s position and generating clouds, these configurations typically require manual intervention and lack calibration. Automating and calibrating this process can be beneficial for the generation of more realistic data, thus improving the quality of the training data.
Objective & Method
The objective of this work is to create a pipeline for the generation of realistic lighting and weather conditions in the Unreal Engine, driven by real-time data. The pipeline should be calibrated using luminescence sensors to recreate the lighting conditions within the Unreal Engine. Thus, the work combines the development of a virtual measuring plugin for luminescence in the Unreal Engine, real and virtual measurements in a Digital Twin environment, and the data-driven generation of Sky Lights and Volumetric Clouds. It is embedded in a deep learning-based approach for drone detection using a combination of synthetic and real-world data, and thus might contribute to bridging the simulation-reality gap.
Work Content
- Generating an Unreal Engine plugin that allows luminescence measurements on surfaces
- Implementing a pipeline for the automatic generation of real data-driven lighting and weather conditions in the Unreal Engine
- Calibrating the pipeline using real and virtual measurement data
What we expect from you:
- currently enrolled as a Master student in computer science, mathematics, optical engineering or similar major
- experience in working with Unreal or Unity (or something comparable)
- interest in deep learning, image processing, and object detection
- basic Python and C++ skills
What we offer:
DLR stands for diversity, appreciation and equality for all people. We promote independent work and the individual development of our employees both personally and professionally. To this end, we offer numerous training and development opportunities. Equal opportunities are of particular importance to us, which is why we want to increase the proportion of women in science and management in particular. Applicants with severe disabilities will be given preference if they are qualified.
Further information:
Starting date: 01.09.2024
Duration of contract: six months
Type of employment: 7 hours per week
Remuneration: depending on qualifications up to 05 TVöD
Duration of contract: six months
Type of employment: 7 hours per week
Remuneration: depending on qualifications up to 05 TVöD
Vacancy-ID: 95199
Contact: Tobias Koch, phone +49 2241 20148 55