Senior Perception Engineer – 3D Computer Vision & Sensor Fusion (human)
NEURA Robotics · Metzingen / Riederich · Posted Oct 30, 2026
Transparency
StaleOnline for 3 weeks and never reposted. The first wave of applicants is in, so a reply is less likely. Apply if it's a strong fit.
- First seen
- Oct 31, 2026
- Offered for
- 21 days
Overview
- Category
- AI & machine learning
- Employment type
- Permanent
- Workplace type
- On-site
- Country
- Germany
- Seniority
- Senior
- Schedule
- Full time
- Salary
- Not listed
Key skills
- Perception pipeline design
- Design and implement end‑to‑end pipelines processing camera, depth and LiDAR data into world models.
- Geometric vision and point cloud algorithms
- Develop stereo matching, registration and model‑based pose estimation algorithms for 3D data.
- Sensor calibration and fusion
- Own calibration, time synchronization and multi‑sensor fusion of cameras, depth sensors and LiDAR.
- State estimation and tracking
- Create filtering and optimization based estimators for robust robot state tracking.
- Real‑time embedded implementation
- Ensure perception pipelines run in real time on embedded Linux hardware.
- Evaluation and benchmarking
- Set up benchmarks, ground‑truth data and metrics to improve precision and robustness.
Key phrases
- classical perception stack
- perception pipelines
- camera, depth and LiDAR data
- geometric vision
- point cloud processing
- stereo matching
- registration
- model-based pose estimation
- sensor calibration
- time synchronization
- multi-sensor fusion
- state estimation
- filtering and optimization methods
- real time on embedded hardware
- evaluation, benchmarks and ground truth
From the posting
Your mission & challenges
Technical direction: You shape the technical direction of our classical perception stack. You define algorithms, architecture and best practices, and drive them into the product.
Perception pipelines: You design and implement perception pipelines for camera, depth and LiDAR data, from raw sensor input to a usable world model.
Geometric vision & point clouds: You develop geometric vision and point cloud processing algorithms, such as stereo matching, registration and model-based pose estimation.
Calibration & fusion: You own sensor calibration, time synchronization and multi-sensor fusion.
State estimation: You develop state estimation and tracking based on filtering and optimization methods.
Real-time performance: You make sure pipelines run in real time on embedded hardware and stay robust under real-world conditions.
Evaluation & benchmarks: You set up evaluation, benchmarks and ground truth, and drive measurable improvements in precision and robustness.
Geometric foundation: You provide the geometric foundation (calibration, 3D data, poses) that the AI team builds on.
Expert & mentor: You act as the go-to expert for classical perception and mentor your colleagues.
What we can look forward to
Education: A Master’s degree or PhD in computer science, robotics, electrical engineering, mathematics, physics or a related field.
Experience: 5+ years of hands-on experience building classical, model-based perception software for robots, machine vision or comparable real-time systems.
3D geometry: Deep knowledge of 3D geometry, camera models and calibration.
Estimation & optimization: A strong background in estimation and optimization, e.g. Kalman filtering, RANSAC or nonlinear least squares.
Sensor physics: A solid understanding of sensor physics and the error sources of cameras, depth sensors and LiDAR.
Track record: Experience shaping technical concepts and taking them to production.
Programming: Excellent modern C++ and good Python skills on Linux.
Libraries & frameworks: Experience with libraries such as OpenCV, PCL or Eigen, and with ROS 2.
Nice to have: Experience with GPU acceleration or embedded platforms.
Languages: Professional English is required. German at B2 to C1 level is a plus.
