NGC Catalog
CLASSIC
Welcome Guest
Isaac ML 3D Pose
Isaac ML 3D Pose
For copy image paths and more information, please view on a desktop device.
Description
Isaac SIM ML Training Workflow - Object 3D pose estimation - this is a Docker container to help Robotics ML engineers to use Isaac Sim to train object 3D pose estimation (included with Isaac SDK) using simulation data and test it in simulation as well as real world. It is targeted at Robotics App Developer and ML Engineer. Required HW is NVIDIA DGX Station i.e. V100 is preferred; but any CUDA capable GPU will do.Please add descriptionPublisher
NVIDIALatest Tag
2020.2Modified
May 1, 2025Compressed Size
16.2 GBMultinode Support
NoMulti-Arch Support
No2020.2 (Latest) Security Scan Results
Linux / amd64
What is 3D Object Pose Estimation Docker
3D Object Pose Estimation docker enable you to train pose estimation for any model completely in simulations, and to test and run the inference in simulations, as well as the real world. Also you can evaluate your model your model using evaluation workflow.
To learn more about the inner workings of pose_cnn_decoder in the Isaac SDK, you can consult the documentation:
Documentation
Learn more from ISAAC Documentation Guide
Hardware Requirement
NVIDIA Pascal GPU or newer. GTX 1080Ti and Titan V are the minimum recommended GPUs.
End User License Agreement
Check your ISAAC License
Technical Support:
Reach us at ISAAC Developer Forum