Persistent predictor apparatus and methods for task switching
First Claim
1. A non-transitory machine-readable storage medium having instructions embodied thereon, the instructions being executable by one or more processors to effectuate training of a learning component configured to operate a robotic apparatus, the instructions configured to, when executed cause the one or more processors to:
- provide a first control output, the first control output configured to cause the robotic apparatus to execute a first physical action;
receive a feature occurrence information comprising feedback via a sensor of the robotic apparatus;
cause production of a predicted output comprising a first vector from the learning component based on an occurrence of a feature in the received feature occurrence information, the predicted output configured to automatically cause the robotic apparatus to execute a second physical action based on one or more components of the first vector which is selected from a first range of values between a first value and a second value;
produce a training input indicative of a target physical action, the training input comprises one or more components of a second vector selected from a second range of values between a third and fourth value;
cause a combination of the training input and the predicted output via a combiner to produce a second control output comprising a third vector, the second control output configured to maintain execution of the first physical action by the robotic apparatus or execute the second physical action based on an analysis of one or more components of the third vector which is selected from a third range of values between the first value and the second value;
provide the second control output to the learning component; and
adapt the learning component in accordance with a discrepancy measure between the second control output and the predicted output;
wherein the adaptation of the learning configuration is configured to cause the learning component to produce the second control output upon the occurrence of the feature during execution of the first action.
2 Assignments
0 Petitions
Accused Products
Abstract
An apparatus and methods for training and/or operating a robotic device to perform a target task autonomously. The target task execution may be configured based on analysis of sensory context by the robot. Target action may comprise execution of two or more mutually exclusive actions for a given context. The robotic device may be operable in accordance with a persistent switching process. For a given sensor input, the switching process may be trained to select one of two or more alternative actions based on a prior action being executed. Switching process operation may comprise assigning priorities to the available tasks based on the sensory context; the task priorities may be modified during training based on input from a trainer. The predicted task priorities may be filtered by a “persistent winner-take-all process configured to switch from a current task to another task based on the priority breaching a switching threshold.
404 Citations
18 Claims
-
1. A non-transitory machine-readable storage medium having instructions embodied thereon, the instructions being executable by one or more processors to effectuate training of a learning component configured to operate a robotic apparatus, the instructions configured to, when executed cause the one or more processors to:
-
provide a first control output, the first control output configured to cause the robotic apparatus to execute a first physical action; receive a feature occurrence information comprising feedback via a sensor of the robotic apparatus; cause production of a predicted output comprising a first vector from the learning component based on an occurrence of a feature in the received feature occurrence information, the predicted output configured to automatically cause the robotic apparatus to execute a second physical action based on one or more components of the first vector which is selected from a first range of values between a first value and a second value; produce a training input indicative of a target physical action, the training input comprises one or more components of a second vector selected from a second range of values between a third and fourth value; cause a combination of the training input and the predicted output via a combiner to produce a second control output comprising a third vector, the second control output configured to maintain execution of the first physical action by the robotic apparatus or execute the second physical action based on an analysis of one or more components of the third vector which is selected from a third range of values between the first value and the second value; provide the second control output to the learning component; and adapt the learning component in accordance with a discrepancy measure between the second control output and the predicted output; wherein the adaptation of the learning configuration is configured to cause the learning component to produce the second control output upon the occurrence of the feature during execution of the first action. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A robotic apparatus comprising:
-
a sensor interface configured to receive sensor data related to an environment of the robotic apparatus; an interface configured to receive a training input; a feature detection component; an output prediction component in operable communication with the feature detection component and the interface, the output prediction component comprising logic configured to, based on a feature detection indication produced by the feature detection component, to produce a first task activation indication and a second task activation indication; and a switching component configured to, responsive to the production of the feature detection indication; evaluate a current active task indication; based on an occurrence of a first sensory input via the sensor interface during execution of a first physical task by the robotic apparatus, and based on the current active task indication corresponding to a third physical task, provide a first task activation output comprising the first task activation indication, the first task action indication being configured to automatically cause the robotic apparatus to execute the third physical task associated therewith; and based on the occurrence of the first sensory input via the sensor interface during execution of a second physical task by the robotic apparatus, and based on the current active task indication corresponding to a fourth physical task, provide a second task activation output comprising the second task activation indication, the second task action indication being configured to automatically cause the robotic apparatus to execute the fourth physical task associated therewith, the fourth physical task differing from the third physical task; wherein; the third and the fourth physical tasks differ from one another so as to produce outcomes different from one another. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18)
-
Specification