NICR-Grasping

This package contains a general definition of a structure for grasp estimation datasets. For detailed listing of documentation see here.

Datasets

Graspnet-1Billion

To use this repository in conjunction with the Graspnet-1Billion dataset the following steps are required:

1. Install the graspnetapi package from our fork. This should only be necessary if you want to use this repository for evaluation.
2. Set the GRASPNET_PATH environment variable or set the path directly in nicr_grasping.utils.paths.py so that your local dataset can be found by this repository.

Publications

This repository was used for the following publications

IROS2022 - On the Importance of Label Encoding and Uncertainty Estimation for Robotic Grasp Detection

For more details see here.

CASE2024 - GraspTrack: Object and Grasp Pose Tracking for Arbitrary Objects

For more details see here.

Changelog

Version 0.2.0 (Mar 27, 2024)

• reworked evaluation pipeline to be more modular
• added visualization for evaluation pipeline
• fixed typehints

Version 0.1.0 (Oct 5, 2023)

• added evaluation pipeline
• added external dependency to dexnet and meshpy (code is directly included as those packages are not actively maintained and not python3 compatible).
• added GraspTrack project readme
• added testcases