Pioneer 3D Scene Reconstruction Engineer

SpAItial AI; Greater London; On-site; GBP 65,000 - 90,000; Full time

Job summary

SpAItial AI in London seeks a Research Engineer to push the quality of 3D reconstruction using state-of-the-art scene representations and reconstruction methods. You will work across the full spectrum, from per-scene optimization to feed-forward models, choosing representations and methods for complex problems and implementing them with strong coding skills.

The role emphasizes hands-on development, rigorous evaluation on benchmarks, and collaboration with researchers to bring methods into

Qualifications

• Bachelor's or Master's in CS/CV/Graphics or equivalent experience.
• Hands-on knowledge of scene representations and reconstruction methods.
• Strong coding skills and ability to turn papers into working code.
• Solid understanding of multi-view geometry and rendering.
• Experience with deep learning experiments and comparisons you can defend.
• PhD or publications in reconstruction are a plus, not required.

Responsibilities

• Advance the quality of 3D reconstruction from ideas to working methods.
• Tackle geometry, appearance, materials and lighting, and dynamic scenes.
• Build and improve reconstruction methods across per-scene optimization to feed-forward.
• Stay at the frontier of scene representations and adopt new ones when advantageous.
• Tune losses, priors and optimization for fidelity, robustness, and efficiency.
• Create rigorous evaluations on public and internal benchmarks to drive decisions.
• Collaborate with research and engineering to productionize successful methods.

Skills

Scene representations

3D reconstruction

Deep learning frameworks

Coding skills

Education

Bachelor's or Master's in CS/CV/Graphics

PhD or publications in reconstruction (nice-to-have)

Tools

Gaussian splats

Radiance fields

Job description

SpAItial AI in London seeks a Research Engineer to push the quality of 3D reconstruction using state-of-the-art scene representations and reconstruction methods. You will work across the full spectrum, from per-scene optimization to feed-forward models, choosing representations and methods for complex problems and implementing them with strong coding skills.

The role emphasizes hands-on development, rigorous evaluation on benchmarks, and collaboration with researchers to bring methods into