Back to results
US 11,455,501 B2Granted
Response based on hierarchical models
AssigneeHewlett-Packard Development Company, L.P.
InventorsThomas Paula, Carlos Leao
Life of this patent
Estimating the term…
Abstract
Examples disclosed herein relate to determining a response based on hierarchical models. In one implementation, a processor applies a first model to an image of an environment to select a second model. The processor applies the selected second model to the image and creates an environmental description representation based on the output of the second model. The processor determines a response based on the environmental description information.
Claims15
A computing system comprising:
- a storage to store a group of image classification models;
- a processor to:
- apply a first model of the group of image classification models to an image to determine first environmental context information associated with a context of an environment depicted in the image;
- select at least one additional model of the group of image classification models to apply to the image based on the first environmental context information;
- apply the at least one additional model to the image to determine second environmental context information;
- determine a response to a query from a user in the environment based on a context representation generated from the first environmental context information and the second environmental context information.
3 more elements
A method, comprising:
- applying, by a processor, a first neural network model of a group of neural network models to an image of an environment to determine initial environmental context information associated with a context of the environment depicted in the image;
- applying, by the processor, a second neural network model of the group of neural network models to the image to generate additional environmental context information, wherein the second neural network model is selected based on the initial environmental context information;
- creating, by the processor, an environmental description representation based on the initial environmental context information and the additional environmental context information;
- and
- determining, by the processor, a response to a query from a user in the environment based on the environmental description representation.
2 more elements
A machine-readable non-transitory storage medium comprising instructions executable by a processor to:
- apply a first machine learning model of hierarchical cascading machine learning models to an image of an environment to determine a first description of the environment;
- applying, by the processor, at least one additional machine learning model of the hierarchical cascading machine learning models to the image to determine a second description of the environment, wherein the at least one additional model machine learning is selected based on the first description of the environment;
- parse at least one of a query or a command from a user in the environment based on the second description of the environment;
- and
- determine a response to the parsed at least of the query or the command.
2 more elements
Dependencies are inferred from the claim text.