A Combined One-Class SVM and Template-Matching Approach for User-Aided Human Fall Detection by Means of Floor Acoustic Features

Table 5

Data used in “Set 3.”

(a) Composition of “Set 3”

Class	Number of occurrences
Human falls	44
Basket	3
Fork	4
Ball	4
Book	3
Bag	4
Chair	4
Human activity	8
Music	14

(b) Templates of “Set 3”

Class	Number of templates
	Clean	Noisy
Basket	52	57
Fork	57	57
Ball	19	55
Book	53	57
Bag	50	56
Chair	89	89
Human activity	11	4
Music	4	11
Total	335	386