A behavior prediction method and device based on sequence state evolution

Abstract

The invention provides a behavior prediction method and device based on sequence state evolution, and the method comprises the steps: obtaining a human body action sequence of a current period; Inputting the human body action sequence into a preset prediction network to obtain a plurality of confidence coefficients of a plurality of prediction actions, and screening out a plurality of candidate prediction actions from the plurality of prediction actions; Inputting the category label corresponding to each candidate prediction action in the plurality of candidate prediction actions and the humanbody action sequence into a generator network of a preset action prediction model to generate a plurality of first candidate action sequences; Splicing the first candidate action sequence corresponding to each candidate prediction action into a human action sequence to generate a plurality of second candidate action sequences; And judging the truth degrees of the plurality of second candidate action sequences according to a preset truth degree judgment model, and determining the target action sequence as a predicted human body action sequence in the next period. Therefore, the performance ofbehavior prediction is improved when the method is used for behavior prediction.

Description

technical field [0001] The present application relates to the technical field of pattern recognition, in particular to a behavior prediction method and device based on sequence state evolution. Background technique [0002] With the rapid development of human behavior understanding, human behavior understanding has aroused widespread interest and has become an important field in computer vision. The rapid development of human behavior analysis has made motion prediction a new field of human motion analysis and has shown its importance in many applications, such as motion video analysis, abnormal behavior detection and automatic driving. Among them, action prediction refers to inferring unfinished actions from partial videos. [0003] It is very challenging to predict the partial actions that have occurred from the observed sequence. Existing action prediction methods are mainly divided into two categories: template matching and classification methods based on temporal featu...

AI Technical Summary

Problems solved by technology

[0006] For this reason, this application proposes a behavior prediction method based on sequence state evolution to solve the technical problem that the behavior prediction method in the prior art cannot solve the ambiguity of some actions, which leads to prediction errors in the process of action prediction

Images