Web-based and Interactive Learning - Recognition Method for a Humanoid Robot

Abstract

In this paper an Object Learning and Recognition method for a Humanoid is presented. This method tries to take advantage of the Cloud Resources, since it is based on image web search in order to build training sets for learn about objects appearance. In case of unavailability of Internet access, the robot would ask human to show the object and take the images from its camera. This way, our method aims to be a flexible and natural Human-Robot Interaction framework and to give as much autonomy as possible to the robot.