Videos reported here show examples of the results of SENSOPAC.
UEDIN Tactile discrimination of liquid viscosities in real time on a Schunk tactile hand. UEDIN investigated ways to actively extract information from sensory data by using appropriate exploratory movements. For active learning, we are interested in determining the optimal action x* to take during test time such that the Mutual Information between sensors y and hidden latent state theta, I(theta; y | x), is maximized. This framework was successfully implemented in the framework of Gaussian Process sensor action modeling and Quadratic Mutual Information (QMI) based action selection (Saal, Ting, & Vijayakumar, 2010a); (Saal, Ting, & Vijayakumar, 2010b). This theoretically well founded formulation was implemented with excellent results on the tactile discrimination of liquid viscosity in real time on a Schunk tactile hand.