基于ST-GCN的形体动作质量评估算法分析

Analysis of Algorithm for Evaluation of Body Movement Quality Based on ST-GCN

阐述一种基于时空图卷积的太极拳动作质量评估框架,通过轻量级OpenPose姿态估计获得训练动作的骨架序列数据,然后使用时空图卷积网络(Spatial Temporal Graph Convolutional Network,STGCN)提取动作的时空特征与标准动作的特征对比,利用孪生网络学习动作的质量分数。

This paper describes a framework for evaluating the quality of Taijiquan movements based on time-space graph convolution. The skeleton sequence data of training movements are obtained through lightweight Open Pose posture estimation, and then the spatial temporal graph convolutional network(ST-GCN) is used to extract the spatiotemporal characteristics of movements and compare them with the characteristics of standard movements, and the twin network is used to learn the quality score of movements.