16.10.2026 aktualisiert
100 % verfügbar
Senior Computer Vision Engineer
Berlin, Deutschland
Nur Remote
PhD in Physics/Optics, University of Kent at CanterburyÜber mich
Computer vision engineer with deep cross-domain experience in building and validating real-world visual systems.
Skills
Computer VisionMedizinische BildgebungC++BildverarbeitungPythonMachine LearningOpenCV3-d ImagingPyTorchDeep Learning
Computer Vision & Image Processing
Extensive experience developing computer vision systems, including object detection, segmentation, image enhancement, feature extraction, optical flow, tracking, image retrieval, and video analysis. Expertise in both classical computer vision and deep learning approaches.
3D Computer Vision, Localization & Mapping
Expertise in camera geometry, calibration, 6-DoF pose estimation, visual localization, Structure from Motion (SfM), 3D reconstruction, point cloud processing, LiDAR, and mapping. Experience developing robust localization and reconstruction systems for real-world environments.
Machine Learning & Applied AI
Development and deployment of machine learning solutions using PyTorch and related frameworks. Experience with CNNs, visual embeddings, representation learning, multimodal processing, model evaluation, and practical integration of AI into production software.
High-Performance C++ & Algorithm Engineering
Advanced C++ development with a strong focus on computational efficiency, memory optimisation, parallel processing, and scalable algorithms. Experience designing and implementing high-performance numerical and geometric algorithms, including independent open-source software libraries.
Python Development & AI Pipelines
Professional Python development for data processing, computer vision, machine learning and automation. Experience building end-to-end processing pipelines, integrating machine learning models and developing practical software tools.
Real-Time & Embedded Vision Systems
Design and optimisation of real-time vision systems, including video processing pipelines, object tracking, low-latency inference and resource-constrained deployments. Experience with mobile, desktop and edge computing environments.
Image Retrieval & Similarity Search
Expertise in local and global visual descriptors, feature matching, embedding-based image retrieval, approximate nearest-neighbour search and visual recognition. Experience with large-scale visual databases and efficient similarity-search systems.
Sensor Fusion, LiDAR & Spatial Data Processing
Experience combining camera, positioning and motion data for localization, tracking and mapping. Expertise in point cloud processing, spatial data structures, trajectory filtering and geometric reconstruction.
Research, Prototyping & Production Engineering
Over 20 years of experience turning complex technical concepts into working products. Strong research and mathematical background, with a proven ability to investigate unfamiliar problems, develop algorithms, implement prototypes and deliver maintainable production software.
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2026
Sprachen
DeutschMutterspracheEnglischMuttersprache
Projekthistorie
AR-based indoor navigation and visual localization. Redesigned and improved a camera-based indoor localization system (6-DoF positioning), improved image retrieval and matching, developed sensor fusion and trajectory filtering, standardized coordinate systems and pose handling, developed algorithms to derive structured navigation maps from LiDAR point clouds, developed privacy-preserving detection and scene analysis, and mentored junior developers.
Retail Computer Vision & Store Analytics. Worked on automatic planogram discovery from monocular shelf images, shelf detection and geometric rectification, real-time shelf change detection, customer tracking using overhead cameras, store layout and shelf geometry reconstruction from 3D scans and point clouds, and technical presentations of deployed systems.
Vision-based mapping and perception for vehicle fleets. Established computer vision effort, designed dashcam data collection infrastructure, developed monocular visual odometry, prototyped lane and pedestrian detection, built a video-based traffic sign mapping system, and contributed to large-scale dashcam video processing and indexing.