🛠 Tech Stack

💼 About This Role

You'll lead research on diffusion-based generative models for 3D content creation at SpAItial. You'll push the boundaries of physically-grounded 3D world modeling for industries like robotics and AR/VR.

🎯 What You'll Do

• Design and develop diffusion-based methods for 3D generation from images and video.
• Build, train, and optimize 3D diffusion models with novel architectures and losses.
• Experiment with 3D representations including point clouds, meshes, and Gaussian Splatting.
• Collaborate to integrate physics-aware priors into diffusion systems.

📋 Requirements

• PhD in computer science, computer vision, or machine learning.
• Top-tier publication record at CVPR, NeurIPS, or SIGGRAPH.
• Hands-on experience training diffusion models like Stable Diffusion or FLUX.
• Proficiency in PyTorch and large-scale model training.

✨ Nice to Have

• Experience with video diffusion backbones such as WAN.
• Solid understanding of 3D processing concepts like camera geometry and depth.

🎁 Benefits & Perks

• 🔬 Cutting-edge research on world models.
• 🌍 Impact across robotics, AR/VR, gaming.
• 🤝 Collaborative and visionary team.
• 🌈 Diverse and inclusive workplace.