Constellations and the unsupervised learning of graphs

In this paper, we propose a novel method for the unsupervised clustering of graphs in the context of the constellation approach to object recognition. Such method is an EM central clustering algorithm which builds prototypical graphs on the basis of fast matching with graph transformations. Our experiments, both with random graphs and in realistic situations (visual localization), show that our prototypes improve the set median graphs and also the prototypes derived from our previous incremental method. We also discuss how the method scales with a growing number of images.

Design of Activated Carbon/Activated Carbon Asymmetric Capacitors

Supercapacitors are energy storage devices that offer a high power density and a low energy density in comparison with batteries. Their limited energy density can be overcome by using asymmetric configuration in mass electrodes, where each electrode works within their maximum available potential window, rendering the maximum voltage output of the system. Such asymmetric capacitors are optimized using the capacitance and the potential stability limits of the electrodes, with the reliability of the design largely depending on the accuracy and the approach taken for the electrochemical characterization. Therefore, the performance could be lower than expected and even the system could break down, if a well thought out procedure is not followed. In this work, a procedure for the development of asymmetric supercapacitors based on activated carbons is detailed. Three activated carbon materials with different textural properties and surface chemistry have been systematically characterized in neutral aqueous electrolyte. The asymmetric configuration of the masses of both electrodes in the supercapacitor has allowed to cover a higher potential window, resulting in an increase of the energy density of the three devices studied when compared with the symmetric systems, and an improved cycle life.