Chip company NVIDIA developed artificial intelligence that creates virtual environments in 3D based on real life videos, making graphic design faster and cheaper

Briefing

Chip company NVIDIA developed artificial intelligence that creates virtual environments in 3D based on real life videos, making graphic design faster and cheaper

December 31, 2018

Briefing

  • Graphics Design AI – Chip company NVIDIA developed new artificial intelligence (AI) that renders real life videos into 3D environments for faster graphic design
  • Benefits – AI can lower costs and speed up process of producing graphics and developing virtual worlds for automotive, gaming or virtual reality applications
  • Training – Conditional generative neural network was trained using real videos of urban environments, taking into account lighting, as well as materials of objects
  • System Components – Used Tesla V100 graphics processing units (GPUs) on AI platform DGX-1 with PyTorch deep learning framework for training AI
  • Public Introduction – Created driving game using 3D environment developed by AI, which was showcased at NeurIPS conference in Montreal, Canada, held on December 2-8, 2018

Accelerator

Sector

Information Technology

Source

Original Publication Date

December 3, 2018

Leave a comment