Research Scientist 3D Diffusion
SpAItialVisit SpAItial website
SpAItial is an AI company building physically grounded world models that generate persistent, explorable 3D worlds from text, images, and panoramas.
London, United Kingdom
Funding history
About SpAItial
SpAItial Ltd operates the Echo model family, a spatial-AI product available through a web app and developer API for generating, editing, sharing, and exporting 3D Gaussian Splat worlds.
Skills
About the Role
You will lead research on diffusion-based generative models for high-quality 3D content. You will develop architectures, losses, and sampling strategies; adapt image and video diffusion backbones; experiment with 3D representations; and improve geometry, fidelity, consistency, and model robustness.
Requirements
- PhD in computer science, computer vision, graphics, machine learning, or a related field
- Top-tier publication record at CVPR, ECCV, ICCV, NeurIPS, or SIGGRAPH
- Strong fundamentals in deep learning, generative modeling, diffusion models, and transformer models
- Experience training diffusion models and image or video model stacks
- Understanding of camera geometry, depth, reconstruction, point clouds, meshes, or Gaussian splats
- Proficiency in Python and PyTorch
- Experience in large-scale model training and optimization
- Ability to run rigorous experiments and ship reliable ML code
Responsibilities
- Design and develop diffusion-based methods for 3D generation
- Build, train, optimize, and evaluate 3D diffusion models
- Research model architectures, losses, and sampling strategies
- Adapt image and video diffusion backbones to 3D generation
- Implement 3D representations including point clouds, meshes, and Gaussian splatting
- Develop training pipelines and loss functions
- Integrate physics-aware priors and world-model capabilities
- Analyze model performance, debug failures, and improve quality and robustness