公开(公告)号：US10949658B2

公开(公告)日：2021-03-16

申请号：US16276493

申请日：2019-02-14

Applicant: WRNCH Inc.

Inventor： Colin J. Brown ,  Andrey Tolstikhin ,  Thomas D. Peters ,  Dongwook Cho ,  Maggie Zhang ,  Paul A. Kruszewski

IPC: G06K9/00 ,  G06N3/08 ,  G06N3/04

Abstract: This disclosure is directed to an activity classifier system, for classifying human activities using 2D skeleton data. The system includes a skeleton preprocessor that transforms the 2D skeleton data into transformed skeleton data, the transformed skeleton data comprising scaled, relative joint positions and relative joint velocities. It also includes a gesture classifier comprising a first recurrent neural network that receives the transformed skeleton data, and is trained to identify the most probable of a plurality of gestures. There is also an action classifier comprising a second recurrent neural network that receives information from the first recurrent neural networks and is trained to identify the most probable of a plurality of actions.

公开(公告)号：US20210161266A1

公开(公告)日：2021-06-03

申请号：US17173978

申请日：2021-02-11

Applicant: WRNCH INC.

Inventor： Colin J. Brown ,  Andrey Tolstikhin ,  Thomas D. Peters ,  Dongwook Cho ,  Maggie Zhang ,  Paul A. Kruszewski

IPC: A45C13/18 ,  A41D27/20 ,  A45C1/02 ,  A45C1/06 ,  A45F5/02

Abstract: An activity classifier system and method that classifies human activities using 2D skeleton data. The system includes a skeleton preprocessor that transforms the 2D skeleton data into transformed skeleton data, the transformed skeleton data comprising scaled, relative joint positions and relative joint velocities. The system also includes a gesture classifier comprising a first recurrent neural network that receives the transformed skeleton data, and is trained to identify the most probable of a plurality of gestures. The system also has an action classifier comprising a second recurrent neural network that receives information from the first recurrent neural networks and is trained to identify the most probable of a plurality of actions.

公开(公告)号：US20190251340A1

公开(公告)日：2019-08-15

申请号：US16276493

申请日：2019-02-14

Applicant: WRNCH Inc.

Inventor： Colin J. Brown ,  Andrey Tolstikhin ,  Thomas D. Peters ,  Dongwook Cho ,  Maggie Zhan ,  Paul A. Kruszewski

IPC: G06K9/00 ,  G06N3/04 ,  G06N3/08

CPC classification number: G06K9/00342 ,  G06K9/00208 ,  G06N3/0454 ,  G06N3/08

Abstract: This disclosure is directed to an activity classifier system, for classifying human activities using 2D skeleton data. The system includes a skeleton preprocessor that transforms the 2D skeleton data into transformed skeleton data, the transformed skeleton data comprising scaled, relative joint positions and relative joint velocities. It also includes a gesture classifier comprising a first recurrent neural network that receives the transformed skeleton data, and is trained to identify the most probable of a plurality of gestures. There is also an action classifier comprising a second recurrent neural network that receives information from the first recurrent neural networks and is trained to identify the most probable of a plurality of actions.