SLAM算法工程师（具身/世界模型方向）-AI数据与安全
Core
Design and optimize visual SLAM/VIO/SfM algorithms for embodied intelligence and world models, focusing on camera pose estimation, trajectory optimization, and 3D structure information.
Role type
Senior IC machine-learning engineer (visual SLAM/embodied AI)
Builds
High-quality trajectory and 3D structure data for long-sequence stability and consistency in embodied systems.
Domain
Robotics, Computer Vision, Embodied AI
Deliverable
production ML models
Required skills
Visual SLAM/VIO/SfM principles, Multi-view geometry (PnP, triangulation, bundle adjustment), Feature-based or direct methods (ORB, DSO), Non-linear optimization (Gauss-Newton, LM), C++, Python
Preferred skills
Hand/body pose estimation, 3D reconstruction (NeRF, Gaussian Splatting), Learning-based SLAM/pose estimation, Long video or large-scale data processing
Technologies
ORB-SLAM, VINS, DSO, COLMAP, NeRF, Gaussian Splatting
