(Fresh)Auto-labelling Algorithm Engineer

Mercedes-Benz Group China Ltd. was founded in Beijing in 2001 and is responsible for the activities of Mercedes-Benz Cars and Vans, Mercedes-Benz Financial Services, Mobility Services, the spare parts business, as well as research and development. With its high-quality products, customer-oriented services, continuous investments, and long-lasting alliances with strong local partners, Mercedes-Benz Group China steadily strives to prove its commitment to its Chinese customers.

(Fresh)Auto-labelling Algorithm Engineer

Objective of job 

• Design and develop MB advanced auto-labelling platform, include 4Dobject, Landmark and BEV auto-labelling functions.
• Architecture design for ground truth training datasets release.
• Auto-labelling pipeline CICD and MLOps.
• Cross function coordinate between auto-labelling pipeline and manual labelling services.
• Define manual labelling software requirement, coordinate the labelling supplier.
• Labelling platform architecture design.
• State-of-the-art auto-labelling algorithm research and study.

Task description

• Design and develop MB advanced auto-labelling platform, include 4Dobject, Landmark and BEV auto-labelling functions.
• Auto-labelling pipeline CICD and MLOps
• Coordinationbetween auto-labelling pipeline and manual labelling services, define manual labelling software requirement, coordinate the labelling supplier.
• Next generation labelling platform architecture design.

Qualification
Master’s degree or above in Computer Science/Robotics/Electronics/Automation and other related majors;
Deep understanding and hands on experience on computer vision, deep learning and robotics.
Solid research background on sensor fusion, 3D object detection, object tracking.
Hands on experience with auto-labelling system or training dataset generation.
Hands on experience with object detection, segmentation based on LiDAR.
Familiar with ROS programming and related tools usage.
Proficient in C++ and Python programming. Good code style is needed.
Hands on experiences on HD mapping automation is a plus.
Hands on experiences on BEV related algorithm development or deployment is a plus.