Research Scientist - 3D Reconstruction (SfM & SLAM)

About the role

SpAItial is pioneering the next generation of World Models, pushing the boundaries of generative AI, computer vision, and realistic simulation. The company is moving beyond 2‑D pixels to build models that natively understand physics and geometry. We are looking for a Research Scientist focused on 3‑D reconstruction to advance methods that recover accurate camera poses and geometry from real‑world imagery.

Key responsibilities

• Design camera‑pose estimators and 3‑D reconstructors, building robust SfM and tracking pipelines for diverse imaging sensors.
• Develop bundle‑adjustment and nonlinear optimisation tools, including non‑perspective camera formulations.
• Integrate and extend state‑of‑the‑art feed‑forward reconstructors such as VGGT, DA3 and Pi3.
• Advance deep multi‑view stereo, learned matching, and monocular depth methods for dense geometry.
• Build evaluation metrics for pose accuracy and reconstruction quality, driving improvements against public and internal benchmarks.
• Scale reconstruction methods to large, diverse real‑world datasets while keeping them reliable and efficient.
• Collaborate with researchers to bring reconstruction advances into production systems.

Required profile

• PhD in computer vision with a research focus on 3‑D reconstruction and publications at top venues (CVPR, ICCV, ECCV, NeurIPS).
• Deep understanding of multi‑view geometry, camera models, epipolar geometry, triangulation and PnP.
• Hands‑on experience with SfM/SLAM systems (e.g., COLMAP, ORB‑SLAM) and nonlinear least‑squares solvers such as Ceres.
• Experience shipping production‑grade 3‑D reconstruction systems is a strong plus.
• Strong Python and PyTorch programming skills.
• Ability to debug failure cases on challenging real‑world data.

Required skills

• Python
• PyTorch
• Ceres Solver
• COLMAP
• ORB‑SLAM
• VGGT
• DA3
• Pi3