Back to results
US 10,474,926 B1Granted
Generating artificial intelligence image processing services
AssigneeAmazon Technologies, Inc.
InventorsLeo Parker Dirac, Vineet Khare, Gurumurthy Swaminathan, Xiong Zhou
Life of this patent
Estimating the term…
Abstract
Features related to systems and methods expediting generation of a machine learning model, such as an image recognition model, are described. Existing machine learning models are analyzed to identify a starting point for creating the new machine learning model. An existing machine learning model can suggest learning parameters (e.g., training parameters or structural features of the model) that can be used to expedite the generating and training process along with training data that can augment the training of the new machine learning model.
Claims19
A computer-implemented method comprising:
- under control of one or more processors,
- receiving, from an electronic communication device, a request for an image model, wherein the request identifies:
- (i) a task for the image model to perform, wherein the task comprises at least one of identifying a location of an object within an image or identifying the object within the image, and
- (ii) training data including a labeled image, wherein the labeled image includes an identification of pixels showing the object;
- identifying, from a library of machine learning models, a first machine learning model and a second machine learning model, based on the task or the object;
- determining a first confidence level that the first machine learning model identifies the pixels showing the object included in the labeled image;
- determining a second confidence level that the second machine learning model identifies the pixels showing the object included in the labeled image;
- determining that the first confidence level is greater than the second confidence level;
- generating a learning parameter for the image model based on a set of labeled image data, the first machine learning model, and the second machine learning model, wherein the learning parameter defines a property of the image model or a property of a training of the image model;
- generating the image model from the first machine learning model, the training data, and the learning parameter;
- and
- storing the image model in the library of machine learning models.
9 more elements
A system comprising:
- one or more computing devices having a processor and a memory, wherein the one or more computing devices execute computer-readable instructions to:
- receive, from an electronic device, a request for an image model that identifies a characteristic of an object shown in a set of reference images;
- determine a first image processing result indicating a first confidence for detecting the characteristic of the object shown in an image included in the set of reference images;
- determine a second image processing result indicating a second confidence for detecting the characteristic of the object shown in the image included in the set of reference images;
- determine that the first confidence is associated with a higher confidence level than the second confidence;
- select, from a library of machine learning models, a first machine learning model based on the request and the first confidence;
- generate the image model from the first machine learning model and the set of reference images;
- and
- store the image model in the library of machine learning models.
6 more elements
A computer-implemented method comprising:
- under control of one or more processors,
- receiving, from an electronic device, a request for an image model that identifies a characteristic of an object shown in a set of reference images;
- determining a first image processing result indicating a first confidence for detecting the characteristic of the object shown in an image included in the set of reference images;
- determining a second image processing result indicating a second confidence for detecting the characteristic of the object shown in the image included in the set of reference images;
- determining that the first confidence is associated with a higher confidence level than the second confidence;
- selecting, from a library of machine learning models, a first machine learning model based on the request and the first confidence;
- generating the image model from the first machine learning model and the set of reference images;
- and
- storing the image model in the library of machine learning models.
6 more elements
Dependencies are inferred from the claim text.