Lead cutting-edge research at Mecka AI as a Neural Objects Researcher, focusing on 3D object generation from video footage. Tackle complex challenges including occlusions, deformable objects, and create simulatable assets.

In this role, you will develop architecture for reconstructing 3D objects from egocentric video. As a pioneering researcher, you will contribute to neural 3D reconstruction, generative refinement, and contact-aware techniques. Collaborate closely to ensure results are transferrable to production while exploring open-ended problems in a data-rich environment.

Key Responsibilities:
• Generate 3D reconstructions from egocentric video inputs
• Implement generative techniques for refining reconstruction quality
• Create assets with physical properties for simulations




• Utilize hand contact information to improve object accuracy
• Prototype research models for practical applications

Requirements:
• Expertise in neural 3D and 4D reconstruction techniques
• Proficient in Python and PyTorch for model training
• Solid foundation in geometry and differentiable rendering
• Robust ability to tackle open-ended research problems
• Experience in generative 3D methods is a plus

Drive advancements in robotics and AI with your expertise in 3D reconstruction and generative modeling at Mecka AI.
#J-18808-Ljbffr

📌 Neural 3D Object Researcher at Mecka AI (Ontario)
🏢 mecka
📍 Ontario