Generate Synthetic Data Using Omniverse Replicator for Perception Models

, Product Manager, NVIDIA

Omniverse Replicator augments costly, laborious human-labeled real-world data, which can be error prone and incomplete, with the ability to create large and diverse physically accurate data tailored to the needs of autonomous vehicle and robotics developers. It also enables generating ground-truth data that's difficult or impossible for humans to label, such as velocity, depth, occluded objects, adverse weather conditions, or tracking the movement of objects across sensors. In this training lab, you'll learn to generate a synthetic dataset from assets and scenes using Omniverse Replicator in a hands-on GPU environment. By participating in this training lab, you’ll learn the following topics: The case for synthetic data How to place a target asset in randomized environments to create a dataset How to turn the dataset into annotated data for training your own model 

 

Prerequisite(s): 

  • Intermediate understanding of Python (including classes, objects, and decorators) 
  • Basic understanding of machine learning and deep learning concepts and pipelines Internet bandwidth sufficient to support the Omniverse client/server stream (performance will vary) 
  • We highly recommend the latest version of Chrome, a mouse, and a second monitor for optimal experience. 

 

Please disregard any reference to "Event Code" for access to training materials. "Event Codes" are only valid during the original live session. Explore more training options offered by the NVIDIA Deep Learning Institute (DLI). Choose from an extensive catalog of self-paced, online courses or instructor-led virtual workshops to help you develop key skills in AI, HPC, graphics & simulation, and more.

活动: GTC Digital Spring
日期: March 2023
行业: 所有行业
话题: Computer Vision - Image Processing
级别: 中级技术
语言: 英语
话题: Computer Vision
所在地: