Nvidia releases FoundationPose on Hugging Face
A post describes the model as estimating and tracking objects’ position and orientation in 3D, with support for unfamiliar objects without fine-tuning.
TLDR
A post announces Nvidia’s FoundationPose release on Hugging Face. It describes a unified model for six-degrees-of-freedom (6-DoF) object pose estimation and tracking—determining an object’s position and orientation in 3D—and says it works on unfamiliar objects without fine-tuning.
Combined views
17.3K
1 Source, first seen ago