System Design of Integrated Intelligent Platform of National Fitness Based on K-Means Clustering

Table 1

F1 value statistical results of end-point detection of different algorithms.

Noise type	Algorithm	Signal-to-noise ratio/dB
		−5	0	5
Airport	Double Threshold Judgment Method	38.68	53.86	66.57
	Cepstrum Distance Method	26.58	35.04	42.78
	Adaptive Algorithm	66.46	75.78	81.22
Babble	Double Threshold Judgment Method	36.69	53.52	65.15
	Cepstrum Distance Method	28.74	46.37	54.71
	Adaptive Algorithm	69.05	77.91	83.01
Car	Double Threshold Judgment Method	38.90	50.96	65.60
	Cepstrum Distance Method	26.94	47.43	51.22
	Adaptive Algorithm	68.54	76.41	80.37
Exhibition	Double Threshold Judgment Method	36.77	53.58	65.54
	Cepstrum Distance Method	29.24	46.01	53.18
	Adaptive Algorithm	66.89	75.36	82.32
Restaurant	Double Threshold Judgment Method	18.05	36.01	44.79
	Cepstrum Distance Method	59.46	68.61	71.31
	Adaptive Algorithm	27.78	40.51	56.55