Video action recognition method and device, electronic equipment and storage medium
An action recognition and video technology, applied in the video field, can solve the problems of high computational complexity and real-time consumption of a lot of resources, and achieve the effect of saving computing resources, reducing operating costs, and improving recognition speed.
Active Publication Date: 2020-08-14
MIGU CO LTD +1
View PDF7 Cites 8 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] Embodiments of the present invention provide a video action recognition method, device, electronic equipment, and storage medium, which are used to solve the defects of high computational complexity, need to occupy a large amount of resources, and low real-time performance of the video action recognition method in the prior art
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0047] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.
[0048] The dual-stream network in the prior art includes two types of video action recognition models, and the two types of video action recognition models are respectively trained based on image information and optical flow information in the video. The video action recognition method based on the dual-stream network extracts image information an...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The embodiment of the invention provides a video action recognition method and device, electronic equipment and a storage medium. The method comprises the following steps of: acquiring a first frame image sequence from a to-be-identified video; generating a first differential image sequence according to the first frame image sequence; extracting features of frame images in the first frame image sequence in time and space to obtain a first video action recognition result; extracting features of differential images in the first differential image sequence in time and space to obtain a second video action recognition result; and acquiring a final video action recognition result of the first frame image sequence according to the first video action recognition result and the second video actionrecognition result. According to the embodiment of the invention, the optical flow information in the existing double-flow network is replaced by the differential image, so that on the premise of notlosing the recognition precision, a large amount of computing resources can be saved, the operation cost is reduced, and the recognition speed of the algorithm is further improved. And the method ismore suitable for scenes with high real-time requirement.
Description
technical field [0001] The present invention relates to the field of video technology, in particular to a video action recognition method, device, electronic equipment and storage medium. Background technique [0002] Video-based human action recognition can be applied in many fields, such as behavior analysis, human-computer interaction, public safety, intelligent monitoring, etc. The emergence of deep convolutional networks has greatly promoted the development of image classification, target recognition and other technologies, so it has also begun to be widely used in action recognition. [0003] Since the video is composed of continuous multi-frame images, the 2D convolutional network in the deep convolutional network cannot obtain time and motion information, so it cannot be used for video action recognition. [0004] A solution in the prior art is to extract motion information from the video, such as optical flow information, and then design a dual-stream network inclu...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/04
CPCG06V40/20G06V20/42G06V20/46G06N3/045
Inventor 赵家成
Owner MIGU CO LTD