Multiple exemplar-based facial image retrieval using independent component analysis.

In this paper, we design a content-based image retrieval system where multiple query examples can be used to indicate the need to retrieve not only images similar to the individual examples, but also those images which actually represent a combination of the content of query images. We propose a scheme for representing content of an image as a combination of features from multiple examples. This scheme is exploited for developing a multiple example-based retrieval engine. We have explored the use of machine learning techniques for generating the most appropriate feature combination scheme for a given class of images. The combination scheme can be used for developing purposive query engines for specialized image databases. Here, we have considered facial image databases. The effectiveness of the image retrieval system is experimentally demonstrated on different databases.     

Tolerance space and molecular similarity

Molecular similarity methods were used in selecting K nearest neighbors and in estimating mutagenicity of 95 aromatic amines and boiling points of a large set of over 2,900 compounds. Similarity is analyzed in terms of the concept of tolerance space. Specifically, the role of non-transitivity of the tolerance relation in estimating properties using similarity methods is examined.     

A Working Memory Workout: How to Expand the Focus of Serial Attention From One to Four Items in 10 Hours or Less.

Five individuals participated in an extensive practice study (10 1-hr sessions, 11,000 trials total) on a self-paced identity-judgment n-back task (n ranging from 1 to 5). Within Session 1, response time increased abruptly by about 300 ms in passing from n = 1 to n > 1, suggesting that the focus of attention can accommodate only a single item (H. Caravan, 1998; B. McElree, 2001). Within Session 10, response time was dramatically reduced and increased linearly with n for n ≤ 4, with a slope of about 30 ms. The data suggest that working memory consists of a focus of attention governed by a limited-capacity search, expandable through practice, and a content-addressable region outside the focus of attention. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Development of an Expert System for Selection of Dam Type on Alluvium Foundations

. A knowledge-based expert system was developed to aid in the selection of the type of dam. The dam type selector expert system (DTSA ES) was designed to determine the type of dam on the alluvium foundations. Detailed expert knowledge is required to estimate the type of dam and to develop an expert system. The DTSA ES utilizes rules of thumb used by an expert for determining the selection of the type of dam. The DTSA ES was developed using a shell program. The expert system was tested on several dam sites in order to validate the decision obtained. The use of this expert system, containing knowledge about the selection of dam type, can be helpful to students, potential owners or contractors in selecting dam types. The current prototype always needs additional parameters for more detailed analyses of new developments. However, the current DTSA ES is designed to include existing information about dam types.     

Evaluation of effect of hydrogen on toughness of Zircaloy-2 by instrumented drop weight impact testing

Hydride-assisted degradation in fracture toughness of Zircaloy-2 was evaluated by carrying out instrumented drop-weight tests on curved Charpy specimens fabricated from virgin pressure tube. Samples were charged to 60 ppm and 225 ppm hydrogen. Ductile-to-brittle-transition behaviour was exhibited by as-received and hydrided samples. The onset of ductile-to-brittle-transition was at about 130 °C for hydrided samples, irrespective of their hydrogen content. Dynamic fracture toughness (K_ID) was estimated based on linear elastic fracture mechanics (LEFM) approach. For fractures occurring after general yielding, the fracture toughness was derived based on equivalent energy criterion. Results are supplemented with fractography. This simple procedure of impact testing appears to be promising for monitoring service-induced degradation in fracture toughness of pressure tubes.     

An atomic model of the outer layer of the bluetongue virus core derived from X-ray crystallography and electron cryomicroscopy

BACKGROUND: Bluetongue virus (BTV), which belongs to the Reoviridae family and orbivirus genus, is a non-enveloped, icosahedral, double-stranded RNA virus. Several protein layers enclose its genome; upon cell entry the outer layer is stripped away leaving a core, the surface of which is composed of VP7. The structure of the trimeric VP7 molecule has previously been determined using X-ray crystallography. The articulated VP7 subunit consists of two domains, one which is largely alpha-helical and the other, smaller domain, is a beta barrel with jelly-roll topology. The relative orientations of these two domains vary in different crystal forms. The structure of VP7 and the organizations of 780 subunits of this molecule in the core of virus is central to the assembly and function of BTV. RESULTS: A 23 A resolution map of the core, determined using electron cryomicroscopy (cryoEM) data, reveals that the 260 trimers of VP7 are organized on a rather precise T = 13 laevo icosahedral lattice, in accordance with the theory of quasi-equivalence. The VP7 layer occupies a shell that is between 260 A and 345 A from the centre of the core. Below this radius (230-260 A) lies the T = 1 layer of 120 molecules of VP3. By fitting the X-ray structure of an individual VP7 trimer onto the cryoEM BTV core structure, we have generated an atomic model of the VP7 layer of BTV. This demonstrates that one of the molecular structures seen in crystals of the isolated VP7 corresponds to the in vivo conformation of the molecule in the core. CONCLUSIONS: The beta-barrel domains of VP7 are external to the core and interact with protein in the outer layer of the mature virion. The lower, alpha-helical domains of VP7 interact with VP3 molecules which form the inner layer of the BTV core. Adjacent VP7 trimer-trimer interactions in the T = 13 layer are mediated principally through well-defined regions in the broader lower domains, to form a structure that conforms well with that expected from the theory of quasi-equivalence with no significant conformational changes within the individual trimers. The VP3 layer determines the particle size and forms a rather smooth surface upon which the two-dimensional lattice of VP7 trimers is laid down.     

PsyCOP-a psychologically motivated connectionist system for objectperception

A connectionist system has been designed for learning and simultaneous recognition of flat industrial objects (based an the concepts of conventional and structured connectionist computing) by integrating the psychological hypotheses with the generalized Hough transform technique. The psychological facts include the evidence of separation of two regions for identification ("what it is") and pose estimation ("where it is"). The system uses the mechanism of selective attention for initial hypotheses generation. A special two-stage training paradigm has been developed for learning the structural relationships between the features and objects and the importance values of the features with respect to the objects. The performance of the system has been demonstrated on real-life data both for single and mixed (overlapped) instances of object categories. The robustness of the system with respect to noise and false alarming has been theoretically investigated.     

Deep learning frameworks to learn prediction and simulation focused control system models

Deep learning (DL) methods have brought world-shattering breakthroughs, especially in computer vision and classification problems. Yet, the design and deployment of DL methods in time series prediction and nonlinear system identification applications still need more progress. In this paper, we present DL frameworks that are developed to provide novel approaches as solutions to the aforementioned engineering problems. The proposed DL frameworks leverage the advantages of autoencoders and long-short term memory network, which are known being data compression and recurrent structures, respectively, to design Deep Neural Networks (DNN) for modeling time series and nonlinear systems with high performance. We provide recommendations on how deep AEs and LSTMs should be utilized to end up with efficient Prediction-focused (Pf) and Simulation-focused (Sf) DNNs for time series and system identification problems. We present systematic learning methods for the DL frameworks that allow straightforward learning of Pf-DNN and Sf-DNN models in detail. To demonstrate the efficiency of the developed DNNs, we present various comparative results conducted on the benchmark and real-world datasets in comparison with their conventional, shallow, and deep neural network counterparts. The results clearly show that the deployment of the proposed DL frameworks results with DNNs that have high accuracy, even with a low dimensional feature vector.

     

Synchronization analysis of coupled Lienard-type oscillators by averaging

Sufficient conditions for the synchronization of coupled Lienard-type oscillators are investigated via averaging technique. The coupling considered here is fixed, nonsymmetric, and nonlinear. Under the assumption that the interconnection topology defines a connected graph, it is shown that the solutions of oscillators converge arbitrarily close to each other, starting from initial conditions arbitrarily far apart, provided that the frequency of oscillations is large enough and the initial phases of oscillators all lie in an open semicircle. It is also shown that the nearly-synchronized oscillations always take place around some fixed magnitude independent of the initial conditions and the coupling functions.