Grasp-Anything-6D

Built upon Grasp-Anything [1], Grasp-Anything-6D [2] is a large-scale dataset for the language-driven 6-DoF grasp detection task with 1M point cloud scenes and more than 200M language-associated 3D grasp poses.

Samples

Language-driven 6-DoF Grasping

We introduce a new language-driven 6-DoF grasping method with a new concept of negative prompt guidance.

Our method enables instruction-following grasping in complex 3D environments.

Robotic Applications

Our method powers real robotic manipulation capability.

References

• [1] An Dinh Vuong, Minh Nhat Vu, Hieu Le, Baoru Huang, Binh Huynh, Thieu Vo, Andreas Kugi, Anh Nguyen. Grasp-anything: Large-scale grasp dataset from foundation models. In ICRA, 2024.

• [2] Toan Nguyen, Minh Nhat Vu, Baoru Huang, An Dinh Vuong, Quan Vuong, Ngan Le, Thieu Vo, Anh Nguyen. Language-Driven 6-DoF Grasp Detection Using Negative Prompt Guidance. In ECCV, 2024.