DYNAMIC IMAGE RECOGNITION MODEL UPDATES

• US 20160163029A1
• Filed: 12/05/2014
• Published: 06/09/2016
• Est. Priority Date: 12/05/2014
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group), Kia Corp.
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Gibbon, David Crawford, Stern, Benjamin J., Liu, Zhu, Zavesky, Eric
• Granted Patent: US 10,115,185 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06K 2207/1012 Special detection of object
• CPC Class Codes: G06T 2207/20208 High dynamic range [HDR] image processing
• CPC Class Codes: G06T 5/92 based on global image properties
• CPC Class Codes: G06V 10/772 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
• CPC Class Codes: G06V 20/10 Terrestrial scenes scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

A method includes receiving first image data at an electronic device, and performing a first image recognition operation on the first image data based on a first image recognition model stored in a memory of the electronic device. The method may include sending an image recognition model update request from the electronic device to a server, in response to determining that a result of the first image recognition operation fails to satisfy a confidence threshold. The method includes receiving image recognition model update information from the server and updating the first image recognition model based on the image recognition model update information to generate a second image recognition model. The method further includes performing a second image recognition operation based on the second image recognition model.

Claims

1. A method comprising:

receiving first image data at an electronic device;

performing, at the electronic device, a first image recognition operation on the first image data based on a first image recognition model stored in a memory of the electronic device;

in response to determining that a result of the first image recognition operation fails to satisfy a confidence threshold, sending an image recognition model update request from the electronic device to a server;

receiving, at the electronic device, image recognition model update information from the server;

updating the first image recognition model based on the image recognition model update information to generate a second image recognition model to be stored in the memory of the electronic device; and

performing, at the electronic device, a second image recognition operation based on the second image recognition model.

2. The method of claim 1, wherein the first image recognition model is a first facial recognition model, and wherein the performing the first image recognition operation includes:

identifying a first face in the first image data based on the first facial recognition model;

determining a confidence value by comparing data associated with the first face to a first set of data that is stored in the memory, wherein the first set of data includes data associated with a first set of faces; and

determining whether the confidence value satisfies the confidence threshold, wherein the confidence value represents a likelihood that the first face matches a particular face of the first set of faces.

3. The method of claim 2, wherein the image recognition model update information received from the server includes data associated with a second set of faces.

4. The method of claim 1, wherein the first image recognition model is a first object recognition model, and wherein the performing the first image recognition operation includes:

identifying a first object in the first image data based on the first object recognition model;

determining a confidence value by comparing data associated with the first object to a first set of data that is stored in the memory, wherein the first set of data includes data associated with a first set of objects; and

determining whether the confidence value satisfies the confidence threshold, wherein the confidence value represents a likelihood that the first object matches a particular object of the first set of objects.

5. The method of claim 4, wherein the image recognition model update information received from the server includes data associated with a second set of objects.

6. The method of claim 1, wherein the second image recognition operation is performed on the first image data based on the second image recognition model.

7. The method of claim 1, further comprising receiving second image data at the electronic device, wherein the second image recognition operation is performed on the second image data based on the second image recognition model.

8. The method of claim 1, further comprising sending location data associated with the first image data from the electronic device to the server, wherein the image recognition model update information is determined based at least in part on the location data.

9. The method of claim 1, wherein the first image recognition model includes a first object recognition model, the method further comprising:

sending the first image data from the electronic device to the server, wherein the first image data includes information associated with a first object that is identified based on the first object recognition model; and

receiving image capture instructions from the server, wherein the image capture instructions provide an indication to the electronic device to capture second image data.

10. The method of claim 9, further comprising:

sending the second image data from the electronic device to the server, wherein the second image data includes information associated with a second object that is identified based on the second object-image recognition model; and

receiving location information from the server, wherein the location information is determined based on the information associated with the first object and based on the information associated with the second object.

11. The method of claim 9, wherein the first image data is captured by a camera of the electronic device based on a first set of image capture parameters, the method further comprising configuring the camera to capture the second image data based on a second set of image capture parameters.

12. The method of claim 11, wherein the image capture instructions prompt a user of the electronic device to adjust the camera prior to capturing the second image data.

13. The method of claim 12, wherein the image capture instructions include instructions to adjust an orientation of the camera, instructions to adjust an orientation of the electronic device, or a combination thereof.

14. The method of claim 1, wherein the confidence threshold is dynamically determined.

15. A computer-readable storage device comprising instructions executable by a processor of an electronic device to perform operations including:

receiving first image data;

performing a first image recognition operation on the first image data based on a first image recognition model;

in response to determining that a result of the first image recognition operation fails to satisfy a confidence threshold, sending an image recognition model update request to a server;

receiving image recognition model update information from the server;

updating the first image recognition model based on the image recognition model update information to generate a second image recognition model; and

performing a second image recognition operation on the first image data or on second image data that is different from the first image data based on the second image recognition model.

16. The computer-readable storage device of claim 15, wherein performing the first image recognition operation includes:

identifying a first face or a first object in the first image data based on the first image recognition model;

determining a confidence value by comparing data associated with the first face or the first object to a first set of data; and

determining whether the confidence value satisfies the confidence threshold, wherein the confidence value represents a likelihood that the first face or the first object matches a particular face or a particular object.

17. The computer-readable storage device of claim 15, the operations further comprising sending location data associated with the first image data to the server, wherein the image recognition model update information is determined based at least in part on the location data.

18. An electronic device comprising:

a camera configured to capture image data;

a processor; and

a memory in communication with the processor, the memory including instructions executable by the processor to perform operations including;

performing a first image recognition operation on first image data captured by the camera, wherein the first image recognition operation is performed based on a first image recognition model stored in the memory;

in response to determining that a result of the first image recognition operation fails to satisfy a confidence threshold, sending an image recognition model update request to a server;

receiving image recognition model update information from the server;

updating the first image recognition model based on the image recognition model update information to generate a second image recognition model; and

performing a second image recognition operation on second image data captured by the camera, wherein the second image recognition operation is performed based on the second image recognition model.

19. The electronic device of claim 18, the operations further comprising modifying image capture parameters of the camera prior to capturing the second image data.

20. The electronic device of claim 18, the operations further comprising:

receiving image capture instructions from the server; and

generating an indication responsive to the image capture instructions, wherein the indication prompts a user to adjust image capture parameters of the camera, to adjust an orientation of the camera, to adjust an orientation of the electronic device, or a combination thereof.