Human body action evaluation method, evaluation device and evaluation system

Abstract

The invention provides a human body action evaluation method, device and system. The evaluation method comprises the following steps: training and deploying a human body action evaluation model; evaluating the human body action of the user by using a human body action evaluation model to obtain a feature score; analyzing, by an emotion analysis model, the emotion when the user views the feature score output by the human body action evaluation model, and constructing an error matching library according to an analysis result; updating the human body action evaluation model by using the error matching library; and evaluating the human body action by using the updated human body action evaluation model. According to the motion evaluation method, a measurement mode of deep learning and sentiment analysis is adopted, a traditional measurement method is combined, and the motion evaluation method which can conduct feature measurement learning by itself, find effective features by itself and can conduct automatic learning and automatic evolution when no motion is seen can be disclosed.

Description

technical field [0001] The application belongs to the technical field of human motion analysis, and in particular relates to a human motion evaluation method, evaluation device and evaluation system. Background technique [0002] With the advancement of deep learning technology and high-performance technology, real-time human pose assessment and behavior understanding has become a research hotspot in the field of computer vision. Some research results in the field of computer vision have been successfully applied to many fields, for example, detecting the fall of the elderly through video. In addition, in sports video analysis, sports teaching and other tasks, human action evaluation and similarity evaluation also have broad application prospects. Human behavior is composed of a set of temporal actions. Effectively judging the gap between standard actions (coach actions) and input actions (student actions) is an important task to measure the similarity between standard acti...

AI Technical Summary

Problems solved by technology

However, some sports actions are not only produced by some momentary actions, but are the result of a combination of actions over a period of time. It is difficult to fully summarize all human action behaviors with artificially designed action feature metrics.

In addition, with these methods, once an unrecognized situation occurs, the recognition cannot be automatically corrected. It is necessary to manually analyze the feature measure of the unrecognized action, and then add the feature measure to the system before the system can recognize this action.

Images