Retrieval of Misplaced Items Using a Mobile Robot via Visual Object Recognition

Abstract

Finding misplaced items inside a home is still a headache for most people, especially the elderly. Aiming at easing the search of these items in home environments, this paper presents a misplaced items retrieval system based on a service robot. First, a natural human interface is developed on the robot, so that it can recognize voice commands from users. Then, an object recognition based on the deep learning Convolutional Neural Networks (CNN) is applied to detect the misplaced items. Next the searching result is announced verbally to the user, meanwhile the location of the item is indicated in the remote user end. Finally, experimental results verify that the proposed system can provide an effective searching service for misplaced items in a home environment.