Automatic classifier profiles from training set metadata
AssigneeInsitu, Inc.
InventorBen Wiltshire
Life of this patent
Estimating the term…
Abstract
A method of identifying one or more objects in an image file that includes receiving an image file at a first classifier, wherein the image file comprises image metadata that identifies characteristics of the image file. The method includes determining whether the first classifier is trained to identify at least one object in the image file based on a first set of training metadata associated with the first classifier, wherein the first set of training metadata defines a set of characteristics for which the first classifier is trained to identify the at least one object. The method includes identifying the one or more objects in the image file using the first classifier responsive to determining that the first classifier is trained to identify the at least one object.
Claims20
A method of identifying one or more objects in an image file, the method comprising:
- extracting training metadata from a first set of training image files, wherein the training metadata comprises training metadata values;
- generating a profile for a first classifier, including the training metadata, with the profile defining characteristics for which the first classifier is trained to identify;
- appending the profile to the first classifier;
- receiving an image file at the first classifier, wherein the image file comprises image metadata that identifies characteristics of the image file;
- determining whether the first classifier is trained to identify at least one object in the image file based on the profile;
- and
- identifying the one or more objects in the image file using the first classifier responsive to determining that the first classifier is trained to identify the at least one object.
A system for identifying one or more objects in an image file, the system comprising:
- a memory;
- and
- processing circuitry coupled with the memory, wherein the processing circuitry is operable to:
- train a first classifier comprising:
- extract training metadata from a first set of training image files, wherein the training metadata comprises training metadata values;
- generate a profile for the first classifier, including the training metadata, with the profile defining characteristics for which the first classifier is trained to identify;
- append the profile to the first classifier;
- after the first classifier is trained, receive an image file at the first classifier, wherein the image file comprises image metadata that identifies characteristics of the image file;
- determine whether the first classifier is trained to identify at least one object in the image file based on the profile;
- and
- identify the one or more objects in the image file using the first classifier responsive to determining that the first classifier is trained to identify the at least one object.
A non-transitory computer-readable storage medium for identifying one or more objects in an image file, the computer-readable storage medium being non-transitory and having computer-readable program code portions stored therein that in response to execution by processing circuitry, cause an apparatus to at least:
- train a first classifier comprising:
- extract training metadata from a first set of training image files;
- generate a profile for the first classifier, including the training metadata, with the profile defining characteristics for which the first classifier is trained to identify;
- append the profile to the first classifier;
- after the first classifier is trained, receive an image file at the first classifier, wherein the image file comprises image metadata that identifies characteristics of the image file;
- determine whether the first classifier is trained to identify at least one object in the image file based on the profile;
- and
- identify the one or more objects in the image file using the first classifier responsive to determining that the first classifier is trained to identify the at least one object.
Dependencies are inferred from the claim text.