Senior Software Engineer, 3d, 4d Reconstruction

NVIDIA NVIDIA · Semiconductors · Shanghai, China +2

Senior Software Engineer role focused on applying deep learning and computer vision to 3D/4D world modeling for autonomous driving products, involving building reconstruction systems, developing evaluation methods, and optimizing neural network performance.

What you'd actually do

  1. Build innovative world reconstruction systems using large geometry models, perception models, Gaussian splatting, diffusion models, and generative world models.
  2. Invent evaluation methods to measure the reconstruction and simulation quality.
  3. Develop tools to visualize and triage evaluation metrics.
  4. Build automated and agentic workflows to boost developer efficiency.
  5. Profile and optimize the performance of neural networks training and deployment.

Skills

Required

  • BS, MS, or PhD degree or equivalent experience in Engineering or Computer Science with a focus on Generative AI, Deep Learning, Computer Vision, Robotics, Computer Graphics, or a related field.
  • 5+ years Hands on experience with structure from motion, dense reconstruction, Gaussian splatting, diffusion models, and/or generative world models.
  • Solid fundamentals in 3D computer vision and deep learning.

Nice to have

  • Experience in autonomous driving: evidence of practical experiments and projects within the autonomous driving domain, showcasing your ability to apply machine learning algorithms to solve sophisticated problems in this field.
  • Published research: particularly around Gaussian splatting, image/video diffusion models, or generative world models demonstrating a deep understanding and contribution to the field.

What the JD emphasized

  • 5+ years Hands on experience with structure from motion, dense reconstruction, Gaussian splatting, diffusion models, and/or generative world models.
  • Solid fundamentals in 3D computer vision and deep learning.

Other signals

  • autonomous driving
  • 3D/4D world modeling
  • generative AI
  • deep learning
  • computer vision
  • simulation