云端算法专家（静态数据产线方向)_XC
Core
Develop multi-frame fusion, cross-view association, and geometry/semantic joint optimization algorithms for lane lines, road boundaries, markings, and static traffic signs to enhance annotation precision, completeness, and spatial consistency.
Role type
Senior IC computer vision algorithm engineer (static data production)
Builds
Automated static road data production pipelines for autonomous driving perception and high-definition mapping
Domain
Autonomous driving, computer vision, 3D reconstruction, high-definition mapping
Deliverable
production ML models
Required skills
Computer vision, 3D geometry, machine learning, multi-view geometry, multi-sensor fusion, point cloud processing, semantic segmentation, object detection, vectorization, camera/LiDAR calibration, spatiotemporal synchronization, Python, C++, deep learning frameworks, algorithm deployment, large-scale data processing
Preferred skills
Static scene automatic annotation, offline perception, high-definition map production, BEV data production, visual foundation models, semi-supervised learning, active learning, model-assisted annotation, complex intersection/long-tail scenario optimization, cloud GPU optimization, distributed inference
Responsibilities
Design complete solutions for actual data problems in static road facility annotation; Leverage offline full-sequence information with high-precision positioning and SLAM to improve automated processing in complex/occluded scenarios; Establish quality standards, evaluation systems, and root cause analysis mechanisms for static data; Drive algorithms from prototype validation to scaled production while optimizing efficiency and stability
Seniority
Senior, hands-on IC