We have investigated the role that different connectivity regimes play in the dynamics of a network of Hodgkin-Huxley neurons by computer simulations. The different connectivity topologies exhibit the following features: random topologies give rise to fast system response yet are unable to produce coherent oscillations in the average activity of the network; on the other hand, regular topologies give rise to coherent oscillations, but in a temporal scale that is not in accordance with fast signal processing. Finally, small-world topologies, which fall between random and regular ones, take advantage of the best features of both, giving rise to fast system response with coherent oscillations.
25 February 2009 in Pattern Analysis and Applications
In this paper, we study the robustness of state-of-the-art automatic fingerprint verification systems against hill climbing and brute-force attacks. We compare the performance of this type of attacks against two different minutiae-based systems, the NIST Fingerprint Image Software 2 (NFIS2) reference system and a Match-on-Card based system. In order to study their success rate, the attacks are analyzed and modified in each scenario. We focus on the influence of initial conditions in hill-climbing attacks, like the number of minutiae in the synthetically generated templates or the performance of each type of modification in the template. We demonstrate how slight modifications in the hill-climbing algorithm lead to very different success rates
Chinese Calligraphy have many different styles. Usually beginners will focus on one style and learn it. However, beginners cannot accurately determine the type of Chinese calligraphy. Especial for the western foreigner. Therefore, it has important significance that using a computer to assist in collecting the calligraphy samples and do the correct classification. In addition, it can help some non-speak Chinese library in the classification and identification with some books. It also improves the efficiency that librarian work for some classification with Chinese Calligraphy books. In our research, we have proposed a new method for dealing with Chinese Calligraphy style recognition. The method is based on using the whole page of the document. We use Convolutional Neural Network (CNN) technique and two datasets of different quality for the experiments, then analyses the results. In those experiments, we made some improvements to the CNN structure with dropout layer. The experimental results show that the full-page documents recognition method proposed can achieve high accuracy with the support of CNN technology, and can effectively identify the five different styles of Chinese calligraphy. Compared with the traditional analysis methods with segmentation, our experimental results show that the method based on the full-page document is feasible, avoiding the cumbersome font segmentation problem. This is more efficient and more accurate.