A Novel Model for Intelligent Pull-Ups Test Based on Key Point Estimation of Human Body and Equipment

Table 4

Comparison of state classification conducted by PEPoseNet and MediaPipe in pull-ups test.

Metrics	Models	States
		Ready or end	Hang	Pull	Achieved	Resume
Accuracy	PEPoseNet	0.998	0.992	0.992	0.994	0.991
	PEPoseNet-D	0.952	0.988	0.986	0.985	0.983
	MediaPipe	0.951	0.986	0.985	0.985	0.982
Precision	PEPoseNet	0.995	0.976	0.989	0.968	0.984
	PEPoseNet-D	0.875	0.954	0.973	0.901	0.957
	MediaPipe	0.871	0.950	0.971	0.904	0.953
Recall	PEPoseNet	0.990	0.978	0.988	0.979	0.981
	PEPoseNet-D	0.721	0.976	0.987	0.972	0.974
	MediaPipe	0.709	0.974	0.985	0.969	0.974
F1 score	PEPoseNet	0.993	0.977	0.988	0.973	0.982
	PEPoseNet-D	0.791	0.965	0.980	0.935	0.966
	MediaPipe	0.782	0.962	0.978	0.935	0.963

The values in bold represent the best data for this metrics.