IP Library › Granted Patent US 11,587,314
Granted Patent B2
US 11,587,314 · App. 16/843,787 · Granted Feb 21, 2023

Intelligent correction of vision deficiency

Inventor: Poorna Kale (Folsom, CA)
Assignee: Micron Technology, Inc.
G06V20/20G06N5/04G06N20/00G06V20/582G06V20/584
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,587,314
App. No.
16/843,787
Granted
Feb 21, 2023
Kind
B2
Abstract

Methods, devices, and computer-readable media for generating color-neutral representations of driving objects are disclosed. In one embodiment, a method is disclosed comprising capturing an image, the image including an object of interest; identifying the object of interest in the image based on identifying one or more colors in the image; associating the object of interest with a known traffic object; identifying a color-neutral representation of the known traffic object; and displaying the color-neutral representation to a user.

Claims (40)

1. A method, comprising:

capturing an image, the image including an object of interest;

determining that a level of color blindness deficiency of a user exceeds a threshold value, the level of color blindness comprising a confidence value output by a classification model in response to the image, the classification model trained using images previously labeled by the user during a training phase, a given label comprising a user detection of an object within a given image, and wherein the classification model is further trained based on the given label successfully identifying a first object in one of the images and on a failure of the user to correctly identify a second object in one of the images;

identifying the object of interest in the image based on identifying one or more colors in the image;

associating the object of interest with a known traffic object;

identifying a color-neutral representation of the known traffic object; and

displaying the color-neutral representation to the user.

2. The method of claim 1 , the capturing the image comprising capturing an image of a roadway while a vehicle is in motion, the capturing performed by a camera sensor installed in the vehicle.

3. The method of claim 1 , the identifying the object of interest in the image based on identifying one or more colors in the image comprising inputting the image into a machine learning (ML) model, the ML model configured to output a type of object of interest.

4. The method of claim 1 , the displaying the color-neutral representation comprising associating icons with each of the one or more colors.

5. The method of claim 1 , the displaying the color-neutral representation comprising associating text labels with each of the one or more colors.

6. The method of claim 5 , the displaying the color-neutral representation further comprising associating a position with each label in the text labels.

7. The method of claim 1 , the displaying the color-neutral representation to the user comprising displaying the color-neutral representation on a display selected from a group consisting of an augmented reality (AR) display, a windshield, an instrument panel, and an infotainment system.

8. A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps of:

capturing an image, the image including an object of interest;

determining that a level of color blindness deficiency of a user exceeds a threshold value, the level of color blindness comprising a confidence value output by a classification model in response to the image, the classification model trained using images previously correctly and incorrectly labeled by the user during a training phase, a given label comprising either a user detection of a given object within a given image or a failure to detect the given object within the given image;

identifying the object of interest in the image based on identifying one or more colors in the image;

associating the object of interest with a known traffic object;

identifying a color-neutral representation of the known traffic object; and

displaying the color-neutral representation to the user.

9. The computer-readable storage medium of claim 8 , the capturing the image comprising capturing an image of a roadway while a vehicle is in motion, the capturing performed by a camera sensor installed in the vehicle.

10. The computer-readable storage medium of claim 8 , the identifying the object of interest in the image based on identifying one or more colors in the image comprising inputting the image into a machine learning (ML) model, the ML model configured to output a type of object of interest.

11. The computer-readable storage medium of claim 8 , the displaying the color-neutral representation comprising associating icons with each of the one or more colors.

12. The computer-readable storage medium of claim 8 , the displaying the color-neutral representation comprising associating text labels and associated positions with each of the one or more colors.

13. The computer-readable storage medium of claim 8 , the displaying the color-neutral representation to the user comprising displaying the color-neutral representation on a display selected from a group consisting of an augmented reality (AR) display, a windshield, an instrument panel, and an infotainment system.

14. An apparatus comprising:

a processor; and

a storage medium for tangibly storing thereon program logic for execution by the processor, the stored program logic comprising logic for:

capturing an image, the image including an object of interest,

determining that a level of color blindness deficiency of a user exceeds a threshold value, the level of color blindness comprising a confidence value output by a classification model in response to the image, the classification model trained using labeled images, the labeled images comprising information about whether the user previously successfully identified objects within at least one given image,

identifying the object of interest in the image based on identifying one or more colors in the image,

associating the object of interest with a known traffic object,

identifying a color-neutral representation of the known traffic object, and

displaying the color-neutral representation to the user.

15. The apparatus of claim 14 , the identifying the object of interest in the image based on identifying one or more colors in the image comprising inputting the image into a machine learning (ML) model, the ML model configured to output a type of object of interest.

16. The apparatus of claim 14 , the displaying the color-neutral representation comprising associating icons with each of the one or more colors.

17. The apparatus of claim 14 , the displaying the color-neutral representation comprising associating text labels and associated positions with each of the one or more colors.

18. The method of claim 1 , wherein the classification model is further trained to determine the level of color blindness based on the given label successfully identifying the first object in one of the images and on the failure of the user to correctly identify the second object in one of the images.

19. The method of claim 1 , wherein the training phase comprises presenting unlabeled images to the user, receiving labels from the user with respect to each of the unlabeled images, determining whether each of the labels successfully identifies an object in one of the images or fails to successfully identify the object in one of the images, and training the classification model using the images labeled by the user, the labels, and information indicative of whether each label is correct or not.

20. The method of claim 1 , wherein the identifying the object of interest comprises determining, based on the level of color blindness of the user that the user object of interest.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 16, 2020
From: KALE, POORNA
To: MICRON TECHNOLOGY, INC.
Reel/Frame 053232/0993 →
Continuity (1)
Related Publication 20210319219A1 · Oct 14, 2021
Cited By (1)
US 12,738,072