Title:
GESTURE RECOGNITION METHOD, APPARATUS, AND DEVICE
Document Type and Number:
WIPO Patent Application WO/2019/023921
Kind Code:
A1
Abstract:
The invention provides a gesture recognition method, relating to the technical field of human-computer interaction. The method comprises: extracting M frame images from a first video segment in a video stream; performing, by means of a deep learning algorithm, gesture recognition on the M frame images to obtain a gesture recognition result corresponding to the first video segment; and performing result fusion on gesture recognition results of continuous N video segments contained in the first video segment, to obtain a fused gesture recognition result. The recognition process described above does not require segmenting and tracking of gestures in a video stream. The method recognizes progressive actions by means of a deep learning algorithm with a high computation speed. The method also fuses progressive actions, thereby increasing the speed of gesture recognition and reducing delays in gesture recognition.
Inventors:
WANG LIANG (CN)
XU SONGCEN (CN)
LIU CHUANJIAN (CN)
HE JUN (CN)
XU SONGCEN (CN)
LIU CHUANJIAN (CN)
HE JUN (CN)
Application Number:
PCT/CN2017/095388
Publication Date:
February 07, 2019
Filing Date:
August 01, 2017
Export Citation:
Assignee:
HUAWEI TECH CO LTD (CN)
International Classes:
G06K9/00
Foreign References:
| CN106991372A | 2017-07-28 | |||
| CN104182772A | 2014-12-03 | |||
| CN106295531A | 2017-01-04 | |||
| US20120056846A1 | 2012-03-08 |
Other References:
See also references of EP 3651055A4
Attorney, Agent or Firm:
BEIJING SAN GAO YONG XIN INTELLECTUAL PROPERTY AGENCY CO., LTD. (CN)
Download PDF: