Continuous interaction learning and detection in real-time

Patent metadata

Field	Value
Publication number	US-2016371604-A1
Application number	US-201615182295-A
Country	US
Kind code	A1
Filing date	Jun 14, 2016
Priority date	Mar 15, 2013
Publication date	Dec 22, 2016
Grant date	—

Abstract

Systems and methods may provide for partitioning a plurality of training samples into a first sequential list of centroids, removing one or more repeating centroids in the first sequential list of centroids to obtain a first reduced list of centroids and generating a set of Hidden Markov Model (HMM) parameters based on the first reduced list of centroids. Additionally, a plurality of detection samples may be partitioned into a second sequential list of centroids, wherein one or more repeating centroids in the second sequential list of centroids may be removed to obtain a second reduced list of centroids. The second reduced list of centroids may be used to determine a match probability for the plurality of detection samples against the set of HMM parameters. In one example, the reduced lists of centroids lack temporal variability.

First claim

1 . (canceled)

2 . An apparatus to process training samples, comprising:
a computing device including a sensor to collect a plurality of training samples selected from human motion samples or human verbal samples, the computing device including: a partition module to partition the plurality of training samples into a sequential list of centroids; a filter to remove one or more repeating centroids in the sequential list of centroids to obtain a reduced list of centroids; and a parameter module to generate a set of parameters based on the reduced list of centroids.

3 . The apparatus of claim 2 , wherein the reduced list of centroids is to lack temporal variability.

4 . The apparatus of claim 2 , wherein the sequential list of centroids is to be associated with a set of clusters having a one-to-one mapping to states in a Hidden Markov Model (HMM).

5 . The apparatus of claim 2 , wherein the partition module includes:
a cluster unit to determine a set of clusters for the plurality of training samples; a centroid unit to determine a set of cluster centroids corresponding to the set of clusters; and an assignment unit to assign each training sample in the plurality of training samples to a centroid in the set of cluster centroids.

6 . A method to process training samples, comprising:
partitioning a plurality of training samples selected from human movement samples or human verbal samples into a sequential list of centroids; removing one or more repeating centroids in the sequential list of centroids to obtain a reduced list of centroids; and generating a set of parameters based on the reduced list of centroids.

7 . The method of claim 6 , wherein the reduced list of centroids lacks temporal variability.

8 . The method of claim 6 , wherein sequential list of centroids is associated with a set of clusters having a one-to-one mapping to states in a Hidden Markov Model (HMM).

9 . The method of claim 6 , wherein partitioning the plurality of training samples includes:
determining a set of clusters for the plurality of training samples; determining a set of cluster centroids corresponding to the set of clusters; and assigning each training sample in the plurality of training samples to a centroid in the set of cluster centroids.

10 . A method to process detection samples, comprising:
partitioning a plurality of detection samples selected from human movement samples or human verbal samples into a sequential list of centroids; removing one or more repeating centroids in the sequential list of centroids to obtain a reduced list of centroids; and using the reduced list of centroids to determine a match probability for the plurality of detection samples against a set of parameters associated with a training session.

11 . The method of claim 10 , wherein the reduced list of centroids lacks temporal variability.

12 . The method of claim 10 , wherein the sequential list of centroids is associated with a set of clusters having a one-to-one mapping to states in a Hidden Markov Model (HMM).

13 . The method of claim 10 , wherein using the list of centroids to determine the match probability includes applying a sliding window to the reduced list of centroids, and wherein the sliding window has a fixed width that equals a number of states in a Hidden Markov Model (HMM).

14 . At least one non-transitory computer readable storage medium comprising a set of instructions which, if executed by a computing device, cause the computing device to:
partition a plurality of detection samples selected from human movement samples and human verbal samples into a sequential list of centroids; remove one or more repeating centroids in the sequential list of centroids to obtain a reduced list of centroids; and use the reduced list of centroids to determine a match probability for the plurality of detection samples against a set of parameters associated with a training session.

15 . The at least one computer readable storage medium of claim 14 , wherein the reduced list of centroids is to lack temporal variability.

16 . The at least one computer readable storage medium of claim 14 , wherein the sequential list of centroids is to be associated with a set of clusters having a one-to-one mapping to states in a Hidden Markov Model (HMM).

17 . The at least one computer readable storage medium of claim 14 , wherein the instructions, if executed, cause a computing device to apply a sliding window to the reduced list of centroids to use the list to determine the match probability, and wherein the sliding window is to have a fixed width that equals a number of states in a Hidden Markov Model (HMM).

Assignees

• Intel Corp

Inventors

• Evans Chuck

Classifications

• G06F18/295

  Markov models or related models, e.g. semi-Markov models; Markov random fields; Networks embedding Markov models · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06F18/24137 Primary

  Distances to cluster centroïds · CPC title

• G06N20/00 Primary

  Machine learning · CPC title