Title:Lower Limbs Gesture Recognition Approach to Control a Medical Treatment BedAuthor(s):Christina Tischler, Klaus Pendl, Erwin Schimbäck, Veronika Putz, Christian Kastl, T. Schlechter, Frederik RunteAbstract:Human machine interaction is showing increasing importance in various areas. In this context a gesture control using machine learning algorithms for a contactless control of a therapy table has been identified as interesting application. Predefined lower limb gestures are performed by an operator, classified by a pocket worn tag, and the results are transferred wirelessly to a remote controller. Two algorithms were compared using a k-nearest neighbor (KNN) and a convolutional neural network (CNN), which are responsible for the classification of the gestures. By using the KNN an accuracy in the range of 75\%--82\% was achieved. Compared to KNN, CNN achieves 89.1\% by applying the categorical classifier and 93.7\% by applying the binary classifier. Simplification of work and convenience in using the therapy table can be achieved by high accuracy and fast response of the control system.Booktitle:Computer Aided Systems Theory -- EUROCAST 2022Publisher:Springer Nature SwitzerlandEditor(s):Moreno-Díaz, Robertoand Pichler, Franzand Quesada-Arencibia, AlexisISBN:978-3-031-25312-6Page Reference:page 318--326, 9 page(s)Publishing:2/2023