Systems, methods, and apparatuses for classifying user activity using temporal combining in a mobile device

• US 8,930,300 B2
• Filed: 01/31/2012
• Issued: 01/06/2015
• Est. Priority Date: 03/31/2011
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Grokop, Leonard Henry, Sarah, Anthony, Nanda, Sanjiv
• Primary Examiner(s): Starks, Wilbert L
• Application Number:
  US13/362,893
• Publication Number:
  US 20120254100A1
• Time in Patent Office: 1,071 Days
• Field of Search: 706/52, 706/45
• US Class Current: 706/52
• CPC Class Codes: A61B 5/1123 Discriminating type of movement, e.g. walking or running A61B5/1116, A61B5/112 take precedence
• CPC Class Codes: G06F 18/25 Fusion techniques
• CPC Class Codes: G06F 2218/12 Classification; Matching
• CPC Class Codes: G06N 7/00 Computing arrangements based on specific mathematical models
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06V 40/25 Recognition of walking or running movements, e.g. gait recognition

Abstract

Components, methods, and apparatuses are provided for determining activity likelihood function values for an activity classification for two or more past epochs based, at least in part, on signals from one or more sensors of a mobile device. A method may comprise, for each of a plurality of activity classifications, determining activity likelihood function values for each of the plurality of activity classifications for two or more past epochs. The activity likelihood function values may be based on signals from one or more sensors of a mobile device. The method may also include combining the activity likelihood function values to determine a likelihood function for an activity classification at a present epoch. The method may also include inferring a present activity of a user co-located with the mobile device to be one of the activity classifications based on the determined likelihood functions for the activity classifications at the present epoch.

Claims

1. A method comprising:

for each of a plurality of activity classifications;

determining activity likelihood function values for each of said plurality of activity classifications for two or more past epochs from simultaneous classifiers based, at least in part, on signals from one or more sensors of a mobile device;

combining said activity likelihood function values to determine a likelihood function for an activity classification at a present epoch;

inferring a present activity of a user co-located with said mobile device to be one of the activity classifications based, at least in part, on said determined likelihood function for said activity classification at said present epoch.

2. The method of claim 1, wherein said activity likelihood function values comprise log-likelihoods.

3. The method of claim 1, wherein inferring said present activity comprises filtering said combined likelihood function values by way of Temporal Voting.

4. The method of claim 1, wherein inferring said present activity comprises filtering said combined likelihood function values by way of Maximum Likelihood filtering.

5. The method of claim 1, wherein inferring said present activity comprises filtering said combined likelihood function values by way of Maximum A Priori filtering.

6. The method of claim 1, wherein inferring said present activity comprises filtering said combined likelihood function values by way of a Finite Impulse Response filter.

7. The method of claim 1, wherein inferring said present activity comprises filtering said combined likelihood function values by way of an Infinite Impulse Response filter.

8. The method of claim 1, wherein each of said plurality of activity classifications is mutually exclusive.

9. The method of claim 1, wherein said one or more sensors comprises at least one accelerometer.

10. The method of claim 9, wherein said one or more sensors comprises said at least one accelerometer in each of three linear dimensions.

11. The method of claim 1, wherein said combining said activity likelihood function values to determine said likelihood function for said activity classification at said present epoch further comprises identifying said activity classification having a highest likelihood function most frequently over said two or more past epochs.

12. An apparatus comprising:

for each of a plurality of activity classifications;

means for determining activity likelihood function values for each of said plurality of activity classifications for two or more past epochs from simultaneous classifiers based, at least in part, on signals from one or more sensors of a mobile device; and

means for combining said activity likelihood function values to determine a likelihood function for an activity classification at a present epoch; and

means for inferring a present activity of a user co-located with said mobile device to be one of said plurality of activity classifications based, at least in part, on said determined likelihood function for said activity classification at said present epoch.

13. The apparatus of claim 12, wherein said means for inferring said present activity comprises at least one of Temporal Voting, Maximum Likelihood filtering, Maximum A Priori filtering, Finite Impulse Response filtering, and Infinite Impulse Response filtering.

14. The apparatus of claim 12, wherein said one or more sensors comprises at least one accelerometer in each of three linear dimensions.

15. An article comprising:

non-transitory storage medium having machine-readable instructions stored thereon which are executable by a processor of a mobile device to;

for each of a plurality of activity classifications;

determine activity likelihood function values for an activity classification for two or more past epochs from simultaneous classifiers based, at least in part, on signals from one or more sensors of said mobile device; and

combine said activity likelihood function values to determine a likelihood function for said activity classification at a present epoch; and

infer a present activity of a user co-located with said mobile device to be one of said activity classifications based, at least in part, on said determined likelihood function value for said activity classification at said present epoch.

16. The article of claim 15, wherein said non-transitory storage medium further includes machine-readable instructions stored thereon which are executable by said processor of said mobile device to infer said present activity of said user co-located with said mobile device using Temporal Voting.

17. The article of claim 15, wherein said non-transitory storage medium further includes machine-readable instructions stored thereon which are executable by said processor of said mobile device to infer said present activity of said user co-located with said mobile device using Maximum Likelihood filtering.

18. The article of claim 15, wherein said non-transitory storage medium further includes machine-readable instructions stored thereon which are executable by said processor of said mobile device to infer said present activity of said user co-located with said mobile device using Maximum A Priori filtering.

19. The article of claim 15, wherein said non-transitory storage medium further includes machine-readable instructions stored thereon which are executable by said processor of said mobile device to infer said present activity of said user co-located with said mobile device using Finite Impulse Response filtering.

20. The article of claim 15, wherein said non-transitory storage medium further includes machine-readable instructions stored thereon which are executable by said processor of said mobile device to infer said present activity of said user co-located with said mobile device using Infinite Impulse Response filtering.

21. A mobile device comprising:

one or more sensors; and

a processor to;

for each of a plurality of activity classifications;

determine activity likelihood function values for said plurality of activity classifications for two or more past epochs from simultaneous classifiers based, at least in part, on signals from said one or more sensors;

combine said activity likelihood function values to determine a likelihood function for an activity classification at a present epoch; and

infer a present activity of a user co-located with said mobile device to be one of said activity classifications based, at least in part, on said determined likelihood function for said activity classification at said present epoch.

22. The mobile device of claim 21, wherein said signals from said one or more sensors comprise accelerometer traces in three linear dimensions.

23. The mobile device of claim 21, wherein said processor to infer said present activity of said user co-located with said mobile device implements Temporal Voting.

24. The mobile device of claim 21, wherein said processor to infer said present activity of said user co-located with said mobile device implements Maximum Likelihood filtering.

25. The mobile device of claim 21, wherein said processor to infer said present activity of said user co-located with said mobile device implements Maximum A Priori filtering.

26. The mobile device of claim 21, wherein said processor to infer said present activity of said user co-located with said mobile device implements Finite Impulse Response filtering.

27. The mobile device of claim 21, wherein said processor to infer said present activity of said user co-located with said mobile device implements Infinite Impulse Response filtering.