Human body behavior prediction method and system based on adaptive graph convolutional adversarial network

Abstract

The invention provides a human body behavior prediction method and system based on an adaptive graph convolutional adversarial network, and the method comprises the following steps: obtaining a humanbody behavior action sequence, and carrying out segmentation according to different observation ratios; inputting segmented action sequences into a trained AGCN-AL network and a local network for behavior prediction respectively; and fusing the prediction results of the local network and the AGCN-AL network to serve as a final behavior prediction result, wherein the AGCN-AL network comprises a feature extraction network provided with a graph adaptive graph convolution network module, and a discriminator and a classifier which are separately connected with the feature extraction network, and the local network comprises a feature extraction network and a classifier which are connected in sequence. According to the method, for human skeleton data, the adaptive graph convolutional adversarialnetwork AGCN-AL and the local network of the adaptive graph convolutional adversarial network AGCN-AL are fused, fusion of prediction results of global information and local information is realized, and the accuracy of human behavior prediction is improved.

Description

technical field [0001] The present disclosure relates to the technical field related to image processing, and specifically relates to a human behavior prediction method and system based on an adaptive graph convolutional adversarial network. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art. [0003] The purpose of behavior prediction is to recognize actions before they end. Behavior prediction has broad application prospects in actual production and life. For example, in monitoring scenarios such as security monitoring and medical monitoring, dangerous events can be detected and prevented as early as possible, and the intelligence level of monitoring can be improved. Behavior prediction can also be applied to areas such as assisted driving, automatic driving, and human-machine collaboration. It can not only provide early warning of dangerous event...

AI Technical Summary

Problems solved by technology

[0005] The inventor found that there are few related algorithms for behavior prediction based on human skeleton data, and the existing methods of behavior prediction based on skeleton data either have poor prediction effect or complex network structure, which is not conducive to practical application

Images