Training Collaborative Robots through User Demonstrations
AssigneeIntel Corporation
InventorsVenkataraman Natarajan, Gagan Acharya
Life of this patent
Estimating the term…
Abstract
The present disclosure provides describes to train a multi policy ML model to control robots in a multi-robot system in collaborating to perform a task. For example, trajectories associated with manipulating an object to perform the collaborative task can be determined and an ML model trained to output control actions for the robots in the multi-robot system to collaborate to complete the task.
Claims21
A computing apparatus comprising:
- a processor;
- and
- a memory storing instructions that, when executed by the processor, configure the apparatus to:
- receive indications of a plurality of images, the plurality of images to be captured by an array of cameras positioned in an environment in which at least two expert demonstrators demonstrate a collaborative task involving at least one object;
- identify, from the plurality of images, a plurality of markers;
- identify, from the plurality of images and the plurality of markers, an initial state of the at least one object;
- determine, from the plurality of images and the plurality of markers, a plurality of trajectories associated with the collaborative task;
- and
- train a machine learning (ML) model to output control states for at least two robots to cooperate to perform the collaborative task involving the at least one object based on the plurality of trajectories and the initial state of the at least one object.
A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computer, cause the computer to:
- receive a plurality of images at a computer, the plurality of images to be captured by an array of cameras positioned in an environment in which at least two expert demonstrators demonstrate a collaborative task involving at least one object;
- identify, from the plurality of images, a plurality of markers;
- identify, from the plurality of images and the plurality of markers, an initial state of the at least one object;
- determine, from the plurality of images and the plurality of markers, a plurality of trajectories associated with the collaborative task;
- and
- train a machine learning (ML) model to output control states for at least two robots to cooperate to perform the collaborative task involving the at least one object based on the plurality of trajectories and the initial state of the at least one object.
An apparatus for training robots, comprising:
- means for receiving a plurality of images at a computer, the plurality of images to be captured by an array of cameras positioned in an environment in which at least two expert demonstrators demonstrate a collaborative task involving at least one object;
- means for identifying, by processing circuitry of the computer from the plurality of images, a plurality of markers;
- means for identifying, by the processing circuitry of the computer from the plurality of images and the plurality of markers, an initial state of the at least one object;
- means for determining, by the processing circuitry of the computer from the plurality of images and the plurality of markers, a plurality of trajectories associated with the collaborative task;
- and
- means for training a machine learning (ML) model to output control states for at least two robots to cooperate to perform the collaborative task involving the at least one object based on the plurality of trajectories and the initial state of the at least one object.
Dependencies are inferred from the claim text.