roym899/pose_and_shape_evaluation

Toolbox to evaluate categorical pose and shape estimation methods

29
/ 100
Experimental

This toolbox helps researchers and engineers assess how well their algorithms can identify an object's 3D orientation (pose) and overall shape, especially when working with distinct categories of objects. It takes in the outputs from your object pose and shape estimation methods and provides standardized metrics to evaluate their accuracy and performance. Robotics researchers, computer vision scientists, and anyone developing perception systems for object manipulation or scene understanding would find this useful.

No commits in the last 6 months.

Use this if you are developing or comparing methods that estimate the 3D pose and shape of categorized objects (e.g., 'cup', 'chair') from sensor data, and you need a standardized way to measure their accuracy.

Not ideal if you are working with unstructured point clouds or general 3D reconstruction without a focus on predefined object categories and their specific pose/shape.

robotics computer-vision 3d-object-recognition perception-systems autonomous-systems
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 6 / 25

How are scores calculated?

Stars

29

Forks

2

Language

Python

License

MIT

Last pushed

Feb 23, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/roym899/pose_and_shape_evaluation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.