Methods and apparatus for learning sensor data patterns for gesture-based input
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.
2 Assignments
0 Petitions
Accused Products
Abstract
Methods and systems for learning, recognition, classification and analysis of real-world cyclic patterns using a model having n oscillators, with primary frequency ω1, ω2, . . . , ωn. The state of the oscillators is evolved over time using sensor observations, which are also used to determine the sensor characteristics, or the sensor observation functions. Once trained, a set of activity detection filters may be used to classify sensor data as being associated with an activity.
-
Citations
20 Claims
-
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 Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. An apparatus comprising a non-transitory computer-readable medium having instructions stored thereon that when executed cause a processor to:
-
receive 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, and (iii) an audio sensor, the sensor data comprising a plurality of sequential sensor measurements; determine a set of activity frequencies associated with the sensor data received during the gesture; iteratively calculate; (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 generate the GD filter based at least in part of the set of activity frequencies and the updated observation function.
-
Specification