Algorithm Engineer – REM

Location: Beijing, China; on-site

Company

Building lightweight 2D vector map systems for intelligent driving through learning-based algorithms and vehicle perception data.

What you will do

• Develop learning-based algorithms that reconstruct structured road vector data from crowdsourced vehicle perception records, REM, and multimodal sensor inputs.
• Model road geometry, semantic features, lane connections, and global road topology using spatial reasoning and graph networks.
• Combine deep learning, graph modeling, and computational geometry to solve complex urban-scene challenges.
• Build scalable pipelines for data aggregation, model validation, map optimization, and incremental map updates.
• Optimize model accuracy, robustness, and generalization under occlusion, variable illumination, and unmarked-road conditions.
• Write maintainable, testable production code in Python or C++, participate in code reviews, and support technical iteration.

Requirements

• Master’s degree or Ph.D. in Computer Science, Electronic Engineering, Robotics, or a related field.
• At least 3 years of algorithm development experience in computer vision, spatial modeling, trajectory mining, or robotics.
• Strong programming and algorithm skills in Python or C/C++, with proficiency in at least one deep learning framework.
• Hands-on experience delivering production-level deep learning or visual perception systems.
• Ability to independently investigate ambiguous technical bottlenecks and deliver practical engineering solutions.
• Fluent written and spoken Mandarin and English required.

Nice to have

• Experience with MapTR, VAD, LaneGAP, TopoNet, image stitching, or vectorization algorithms.
• Experience aggregating mass vehicle trajectories and processing crowdsourced perception data.
• Exposure to GIS, computational geometry, SLAM, or lightweight vector map development for ADAS.
• Understanding of CNNs, GNNs, Transformers, object detection, semantic segmentation, or generative AI.
• Experience migrating academic research algorithms into mass-production pipelines.

Culture & Benefits

• Rapid iteration is a core team value.
• Work focuses on fully automated map production pipelines.
• Collaboration includes code reviews and technical iteration.