Back to results
US 11,541,534 B2Granted
Method and system for object grasping
AssigneeIntrinsic Innovation LLC
InventorsLi Yang Ku, Michael Stark, Ahmad Humayun, Nan Rong, Bhaskara Marthi
Life of this patent
Estimating the term…
Abstract
A method for object grasping, including: determining features of a scene; determining candidate grasp locations; determining a set of candidate grasp proposals for the candidate grasp locations; optionally modifying a candidate grasp proposal of the set; determining grasp scores associated with the candidate grasp proposals; selecting a set of final grasp proposals based on the grasp scores; and executing a grasp proposal from the set of final grasp proposals.
Claims24
A method, comprising:
- determining a set of pairs of keypoints and object component identifiers for each of a plurality of objects in a scene, each pair of the set of pairs comprising a respective keypoint and a corresponding object component identifier;
- determining a set of candidate grasp locations, using the sets of pairs, wherein each candidate grasp location of the set of candidate grasp locations is associated with a respective occlusion score determined by a machine learning model trained on images labeled with predetermined occlusion scores for locations in the images;
- determining a set of candidate grasp proposals for the set of candidate grasp locations;
- scoring each of the candidate grasp proposals with a respective score;
- selecting a candidate grasp proposal from the set of candidate grasp proposals as a final grasp proposal based on the respective scores;
- and
- with a robot, executing the final grasp proposal.
4 more elements
A system comprising:
- one or more processors;
- and
- one or more non-transitory machine-readable storage medium storing instructions that are executable by the one or more processors to cause performance of operations comprising:
- determining a set of pairs of keypoints and object component identifiers for each of a plurality of objects in a scene, each pair of the set of pairs comprising a respective keypoint and a corresponding object component identifier;
- determining a set of candidate grasp locations, using the sets of pairs, wherein each candidate grasp location of the set of candidate grasp locations is associated with a respective occlusion score determined by a machine learning model trained on images labeled with predetermined occlusion scores for locations in the images;
- determining a set of candidate grasp proposals for the set of candidate grasp locations;
- scoring each of the candidate grasp proposals with a respective score;
- selecting a candidate grasp proposal from the set of candidate grasp proposals as a final grasp proposal based on the respective scores;
- and
- with a robot, executing the final grasp proposal.
7 more elements
Dependencies are inferred from the claim text.