Computer scientist for traffic area segmentation

Home-Office ArcGIS Computer Vision Deep Learning Forschung GIS Machine Learning

+2 Top

The Remote Sensing Technology Institute is a DLR institute with sites in Oberpfaffenhofen near Munich, Berlin-Adlershof and Neustrelitz in Mecklenburg-Western Pomerania. Together with the German Remote Sensing Data Center, the institute forms the Earth Observation Center EOC, the centre of excellence for earth observation in Germany.

what awaits you

Remote sensing, with its various sensors and platforms, is a valuable data source for traffic research. Entire cities and regions can be captured on a large scale and analyzed with respect to traffic-related questions. The Institute for Remote Sensing Methodology regularly acquires aerial imagery using the aircraft and helicopters of the DLR research fleet, as well as institute-owned camera systems. In addition, the institute has access to high-resolution satellite imagery. To make optimal use of these sensor systems, methods and algorithms are developed for the automatic extraction of traffic objects and traffic areas. These innovative algorithms play a role, for example, in the development of highly accurate, de-tailed maps for automated driving or in improving micro- and macroscopic traffic models. Novel deep learning algorithms achieve very promising results, which can be further improved and adapted to the respective task.

your tasks

• Further development of deep learning algorithms (AI methods) for application on high-resolution aerial and satellite data to capture traffic areas, including their functions (e.g., roads, access routes, bicycle paths, etc.)
• Development of a pre-operational software processor, including AI algorithms, for large-scale mapping of traffic areas
• Validation of results using independent datasets and accuracy assessments of the developed methods
• Collaboration with project partners to utilize the data for addressing traffic science-related questions
• Scientific publication of results and presentation at national and international conferences

• Completed academic university degree (Diploma/Master's) in computer science, machine learning, or a comparable field of study
• Advanced programming skills in Python and experience with deep learning frameworks (especially PyTorch)
• Practical experience with state-of-the-art computer vision and deep learning models such as CNNs and transformers
• Experience in applying AI methods and optimizing performance with respect to accuracy and processing speed
• Ability to work collaboratively in an interdisciplinary team, strong problem-solving, communication, and presentation skills
• Good English skills (B2 level or higher)
• Experience working with remote sensing data and GIS software (e.g., ArcGIS, QGIS) is an advantage