×

Methods and apparatus for learning sensor data patterns for gesture-based input

  • US 9,582,080 B1
  • Filed: 06/25/2015
  • Issued: 02/28/2017
  • Est. Priority Date: 06/25/2014
  • Status: Active Grant
First Claim
Patent Images

1. A method for generating a gesture-detection (GD) filter comprising:

  • receiving sensor data captured by at least one sensor during a gesture, the at least one sensor selected from the group consisting of (i) a motion sensor that is attached to a user, (ii) a motion sensor that is attached to a user equipment, (iii) an audio sensor, and (iv) an electrode array attached to a head of the user, the sensor data comprising a plurality of sequential sensor measurements;

    determining a set of activity frequencies associated with the sensor data received during the gesture;

    iteratively calculating;

    (i) a prediction error based on a difference between at least one of the plurality of sequential sensor measurements and an observation prediction ĥ

    t from an observation model, the observation model operating on phases in a set of phases associated with the set of activity frequencies;

    (ii) an updated set of estimated activity phases based on the prediction error and the set of activity frequencies;

    (iii) an updated observation function of the observation model based on the prediction error; and

    generating the GD filter based at least in part of the set of activity frequencies and the updated observation function.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×