Bernhard Jaeger
Co-Founder at KE:SAI
Tübingen, Baden-Württemberg, Germany
Summary
🤩
Rockstar🎓
Top SchoolBernhard Jaeger is a doctoral researcher and software engineer specializing in embodied intelligence for autonomous driving, with about a decade of experience spanning computer vision, machine learning, robotics and reinforcement learning. He builds and evaluates end-to-end driving systems in closed-loop on public benchmarks, combining research rigor from IMPRS-IS and the Max Planck Institute with hands-on engineering. His background includes a Games Engineering BSc from TUM and industry experience developing graphics for luxury car headlamps, which informs his strength in simulation and sensor-rich perception pipelines. On GitHub he has contributed practical fixes and performance improvements to the well-known TransFuser project, focusing on LiDAR voxelization and waypoint controllers to make transformer-based sensor fusion more reliable. Based in Tübingen, he brings a pragmatic, system-level perspective that bridges academic publication and deployable model engineering.
10 years of coding experience
2 years of employment as a software developer
Master, Informatics, Master, Informatics at University of TübingenBachelor of Science - BS, Informatik: Games Engineering, Bachelor of Science - BS, Informatik: Games Engineering at Technical University of Munich
Stackoverflow
3,573reputation
776kreached
5answers
5questions
Github Skills (17)
imitation-learning10
pytorch10
python10
model-optimization10
sensor-fusion10
autonomous-driving10
transformers9
computer-vision9
configuration-management8
linker6
boost6
keras6
anaconda6
visual-studio6
point-cloud-library6
Programming languages (5)
Github contributions (5)
autonomousvision/transfuser
Jul 2021 - Aug 2022
[PAMI'23] TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Driving; [CVPR'21] Multi-Modal Fusion Transformer for End-to-End Autonomous Driving
Role in this project:
ML Engineer
ML EngineerContributions:24 commits, 1 PR, 22 pushes in 1 year 2 months
Contributions summary:Bernhard primarily contributed to the autonomous driving project by addressing bugs and optimizing the TransFuser model, specifically relating to LiDAR processing and waypoint control. Their work includes fixing voxelization and LiDAR-related issues, improving waypoint controller performance, and integrating best-performing model weights. The user also made modifications to the model and configuration files, indicating a focus on model fine-tuning and performance improvements. These changes reflect an iterative development process to enhance the model's accuracy and reliability.
autonomous-drivingmultimodalsensor-fusionimitation-learningtransformers
Kait0/carla-roach
Aug 2021 - May 2022
Contributions:8 commits, 5 pushes in 8 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.