PHYSICAL MOVEMENT ANALYSIS

• US 20190325780A1
• Filed: 04/18/2019
• Published: 10/24/2019
• Est. Priority Date: 04/19/2018
• Status: Active Grant

• Current Assignee: SSAM Sports Incorporated
• Original Assignee: SSAM Sports Incorporated
• Inventors: KIM, Sang J.
• Granted Patent: US 11,386,806 B2
• Time in Patent Office: Days
• CPC Class Codes: A63B 2071/065 Visualisation of specific exercise parameters
• CPC Class Codes: A63B 2220/803 Motion sensors
• CPC Class Codes: A63B 2220/836 Sensors arranged on the body of the user
• CPC Class Codes: A63B 71/0622 Visual, audio or audio-visual systems for entertaining, instructing or motivating the user
• CPC Class Codes: G06F 3/011 Arrangements for interaction with the human body, e.g. for user immersion in virtual reality blind teaching G09B21/00
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G09B 19/003 Repetitive work cycles; Sequence of movements
• CPC Class Codes: G09B 19/0038 Sports

Abstract

A processing device receives three dimensional (3D) motion capture data corresponding to a subject user performing a physical activity and receives attribute data associated with the subject user. The processing device determines a personalized reference data set for the subject user based on 3D motion capture data associated with a group of users performing the physical activity, wherein each user from the group of users shares at least a portion of the first attribute data with the subject user. The processing device provides the personalized reference data set as an input to a trained machine learning model and obtains an output of the trained machine learning model, wherein the output comprises a recommendation for the subject user pertaining to improvement of the physical activity.

Claims

1. A method comprising:

receiving three dimensional (3D) motion capture data corresponding to a subject user performing a physical activity;

receiving first attribute data associated with the subject user;

determining, by a processing device, a personalized reference data set for the subject user based on 3D motion capture data associated with a group of users performing the physical activity, wherein each user from the group of users shares at least a portion of the first attribute data with the subject user;

providing the personalized reference data set as an input to a trained machine learning model; and

obtaining an output of the trained machine learning model, wherein the output comprises a recommendation for the subject user pertaining to improvement of the physical activity.

2. The method of claim 1, further comprising:

capturing the 3D motion capture data using a plurality of motion capture sensors affixed to one or more body parts of the subject user while the subject user is performing the physical activity.

3. The method of claim 2, wherein the 3D motion capture data comprises one or more of positional data, rotational data, or acceleration data measured by the plurality of motion capture sensors.

4. The method of claim 1, wherein the first attribute data comprises physical characteristic data associated with the subject user, the physical characteristic data comprising at least one of an age, height, weight, or gender of the subject user.

5. The method of claim 1, wherein the first attribute data comprises at least one of range of motion data or functional movement data associated with the subject user.

6. The method of claim 1, wherein determining the personalized reference data set comprises comparing the first attribute data associated with the subject user to second attribute data associated with a plurality of reference data set candidates and identifying the group of users from the plurality of reference data set candidates based on a correlation of the first attribute data with the second attributed data.

7. The method of claim 1, wherein, when executed, the trained machine learning model is configured to:

identify, from the personalized reference data set, a first subset of the group of users for which a corresponding proficiency ranking is greater than a first threshold;

identify, from the personalized reference data set, a second subset of the group of users for which a corresponding proficiency ranking is less than a second threshold; and

determine at least one difference between 3D motion capture data associated with the first subset and 3D motion capture data associated with the second subset, the at least one difference corresponding to a body movement associated with performing the physical activity.

8. The method of claim 7, wherein the recommendation for the subject user pertaining to improvement of the physical activity is based at least in part on the at least one difference.

9. The method of claim 1, wherein the recommendation for the subject user comprises an indication of one or more sub-optimal body movements associated with the subject user performing the physical activity.

10. The method of claim 9, wherein the recommendation for the subject user further comprises an indication of at least one of a practice drill associated with the one or more sub-optimal body movements, or an image or video demonstrating a correct version of the one or more sub-optimal body movements.

11. The method of claim 1 wherein the trained machine learning model is trained using a training data set, the training data set comprising examples of 3D motion capture data associated with users performing the physical activity as a training input and proficiency rankings that indicate how well each of the users performs the physical activity as a target output.

12. A system comprising:

a memory device storing instructions;

a processing device coupled to the memory device, the processing device to execute the instructions to;

receive three dimensional (3D) motion capture data corresponding to a subject user performing a physical activity;

receive first attribute data associated with the subject user;

determine a personalized reference data set for the subject user based on 3D motion capture data associated with a group of users performing the physical activity, wherein each user from the group of users shares at least a portion of the first attribute data with the subject user;

provide the personalized reference data set as an input to a trained machine learning model; and

obtain an output of the trained machine learning model, wherein the output comprises a recommendation for the subject user pertaining to improvement of the physical activity.

13. The system of claim 12, wherein the processing device to execute the instructions further to:

capture the 3D motion capture data using a plurality of motion capture sensors affixed to one or more body parts of the subject user while the subject user is performing the physical activity, wherein the 3D motion capture data comprises one or more of positional data, rotational data, or acceleration data measured by the plurality of motion capture sensors.

14. The system of claim 12, wherein the first attribute data comprises at least one of physical characteristic data, range of motion data, or functional movement data associated with the subject user, the physical characteristic data comprising at least one of an age, height, weight, or gender of the subject user.

15. The system of claim 12, wherein to determine the personalized reference data set, the processing device to execute the instructions to compare the first attribute data associated with the subject user to second attribute data associated with a plurality of reference data set candidates and identify the group of users from the plurality of reference data set candidates based on a correlation of the first attribute data with the second attributed data.

16. The system of claim 12, wherein, when executed, the trained machine learning model is configured to:

identify, from the personalized reference data set, a first subset of the group of users for which a corresponding proficiency ranking is greater than a first threshold;

identify, from the personalized reference data set, a second subset of the group of users for which a corresponding proficiency ranking is less than a second threshold; and

determine at least one difference between 3D motion capture data associated with the first subset and 3D motion capture data associated with the second subset, the at least one difference corresponding to a body movement associated with performing the physical activity, wherein the recommendation for the subject user pertaining to improvement of the physical activity is based at least in part on the at least one difference.

17. The system of claim 12, wherein the recommendation for the subject user comprises an indication of one or more sub-optimal body movements associated with the subject user performing the physical activity and an indication of at least one of a practice drill associated with the one or more sub-optimal body movements, or an image or video demonstrating a correct version of the one or more sub-optimal body movements.

18. A non-transitory computer-readable storage medium storing instructions that, when executed by a processing device, cause the processing device to:

receive three dimensional (3D) motion capture data corresponding to a subject user performing a physical activity;

receive first attribute data associated with the subject user;

determine a personalized reference data set for the subject user based on 3D motion capture data associated with a group of users performing the physical activity, wherein each user from the group of users shares at least a portion of the first attribute data with the subject user;

provide the personalized reference data set as an input to a trained machine learning model; and

obtain an output of the trained machine learning model, wherein the output comprises a recommendation for the subject user pertaining to improvement of the physical activity.

19. The non-transitory computer-readable storage medium of claim 18, wherein the instructions further cause the processing device to:

capture the 3D motion capture data using a plurality of motion capture sensors affixed to one or more body parts of the subject user while the subject user is performing the physical activity, wherein the 3D motion capture data comprises one or more of positional data, rotational data, or acceleration data measured by the plurality of motion capture sensors, and wherein the first attribute data comprises at least one of physical characteristic data, range of motion data, or functional movement data associated with the subject user, the physical characteristic data comprising at least one of an age, height, weight, or gender of the subject user.

20. The non-transitory computer-readable storage medium of claim 18, wherein the recommendation for the subject user comprises an indication of one or more sub-optimal body movements associated with the subject user performing the physical activity and an indication of at least one of a practice drill associated with the one or more sub-optimal body movements, or an image or video demonstrating a correct version of the one or more sub-optimal body movements.