Human body behavior prediction method based on knowledge distillation
A prediction method and knowledge technology, applied in the field of human behavior prediction based on knowledge distillation, can solve the problems of low accuracy of human behavior prediction and increased difficulty of human behavior prediction, and achieve faster operation speed, reduced calculation amount, and accurate experimental results. Effect
Active Publication Date: 2019-11-22
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF8 Cites 7 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] In view of the increasing difficulty of human behavior prediction in the prior art, which causes the problem of low accuracy of human behavior prediction, the present invention provides a human behavior prediction method based on knowledge distillation. The combination of the teacher network and the student network that predicts behavior from some videos can greatly reduce the amount of calculation to a certain extent, speed up the calculation, and obtain accurate experimental results.
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
[0029] All features disclosed in this specification, or steps in all methods or processes disclosed, may be combined in any manner, except for mutually exclusive features and / or steps.
[0030] Combine below Figure 1-Figure 4 The present invention will be described in detail.
[0031] The present invention can be established on the basis of monitoring cameras in public places and electronic eye monitoring systems on traffic roads, etc., and can be applied.
[0032] A human behavior prediction method based on knowledge distillation, comprising the following steps:
[0033] Step 1: Divide the videos in the known human behavior data set into a complete video and a partial video one by one, and perform the operation of extracting video image frames respectively, and then perform data processing on the obtained video image frames, and observe according to the partial The video is divided into a complete video image frame and a partial video image frame set;
[0034] Step 2: Con...
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 invention discloses a human body behavior prediction method based on knowledge distillation, belongs to the field of human body behavior prediction, and particularly relates to a knowledge distillation-based human body behavior prediction method. The problem that in the prior art, the human body behavior prediction accuracy is low due to the fact that the human body behavior prediction difficulty is increased is solved. The technical scheme of the invention is as follows: tdividing videos in the known human body behavior data set into a complete video and a partial video one by one; carrying out video image frame extraction operation on the complete video and the partial video respectively; carrying out data processing on the obtained video image frame; dividing a video into a completevideo image frame and a partial video image frame set according to different partial observation rates, then introducing a teacher network and a student network, and finally, learning the teacher network and the student network by introducing knowledge distillation, so as to identify and predicting behaviors of partial videos. According to the invention, the accuracy of human body behavior prediction can be effectively improved.
Description
technical field [0001] The invention relates to the field of human behavior prediction, in particular to a human behavior prediction method based on knowledge distillation. Background technique [0002] In recent years, with the development of computer vision, human behavior recognition has become a very important research topic, and has been widely used in many fields, such as video surveillance, video understanding, video retrieval and so on. Simply put, the goal of human behavior recognition is the process of accurately classifying a specific behavior segment in a given video. But in daily life, when a traffic accident or a terrorist attack occurs, it is undoubtedly useless to wait for the execution of the behavior to perform human behavior recognition. Therefore, it is necessary to respond immediately before these behaviors occur, for example, predicting the possible serious consequences of drunk driving with a high probability before a traffic accident, or various terr...
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 IPC IPC(8): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06V40/20G06V20/46G06V20/40G06N3/044G06N3/045Y02D10/00
Inventor程建高银星汪雯苏炎洲白海伟
OwnerUNIV OF ELECTRONICS SCI & TECH OF CHINA