Computer Vision Engineer

Computer Vision Engineer – Neural Reconstruction & Generation

About the Role

Join our team building 3D reconstruction and generative AI for sensor simulation. You'll turn real-world sensor data into high-quality static and dynamic 3D scenes that help close the sim-to-real gap. Your work directly shapes how next-generation AI systems are trained and validated.

What You'll Do

• Develop 3D reconstruction and neural rendering methods, including 3D Gaussian Splatting, for large-scale, dynamic driving scenes.

• Apply and adapt video diffusion models to visual generation tasks.

• Process and fuse camera, LiDAR and other sensor data.

• Build sensor calibration and camera/video data tools.

• Define realism metrics, and turn research into reliable, scalable pipelines.

Required Qualifications

• MS or PhD in Computer Vision, Machine Learning, Graphics, Robotics or a related field, or equivalent experience.

• Strong 3D vision fundamentals: camera models, multi-view geometry, depth estimation and rendering.

• Deep experience in either 3D Gaussian Splatting or other neural rendering methods (including dynamic scenes), or video diffusion models, with working knowledge of the other.

• Strong Python and PyTorch skills.

Preferred Qualifications

• Experience with LiDAR and other sensor data, and with multi-sensor calibration.

• C++ and CUDA.

• Strong linear algebra and optimization skills.

• Camera and video data processing experience.

• Large-scale real-world datasets, SLAM or SfM, and top-venue publications.