Photonic-aware neural networks

Neural Computing and Applications - Photonics-based neural networks promise to outperform electronic counterparts, accelerating neural network computations while reducing power consumption and...     

An empirical evaluation of active learning strategies for profile elicitation in a conversational recommender system

Journal of Intelligent Information Systems - Conversational Recommender Systems have received widespread attention in both research and practice. They assist people in finding relevant and...     

Prototyping and modeling of the centrifugal casting process for paraffin waxes

Centrifugal casting is a technology used for manufacturing hybrid rocket paraffin grains. This technology helps avoiding voids formation inside the solid paraffin as it cools. Voids are formed because of air bubbles being entrapped while pouring and because the liquid wax shrinks by 17–19% upon cooling. In this work, the centrifugal casting process for the manufacturing of paraffin cylinders was prototyped at two different scales considering critical casting issues. The effects of process parameters (rotational speed, melt temperature, and flow rate) on the tensile properties of the manufactured grains were analyzed. The results of the optimization conducted at the lower scale (2.5?kg) were up scaled to manufacture 25?kg grains. The resulting mechanical properties complied with the design specifications, and they were better than those characterized from the gravity cast wax. A numerical model of growth and dissolution of bubbles during the process was then developed to predict the quality of the castings. The numerical results showed how increasing the mold rotational speed up to 1800?rpm reduced the removal time. However, compared to grains solidification time, the predicted removal times were much shorter, proving the advantage of centrifugal casting in counteracting voids formation.     

Anticipating Cutoff Diameters in Deterministic Lateral Displacement (DLD) Microfluidic Devices for an Optimized Particle Separation

Deterministic lateral displacement (DLD) devices enable to separate nanometer to micrometer‐sized particles around a cutoff diameter, during their transport through a microfluidic channel with slanted rows of pillars. In order to design appropriate DLD geometries for specific separation sizes, robust models are required to anticipate the value of the cutoff diameter. So far, the proposed models result in a single cutoff diameter for a given DLD geometry. This paper shows that the cutoff diameter actually varies along the DLD channel, especially in narrow pillar arrays. Experimental and numerical results reveal that the variation of the cutoff diameter is induced by boundary effects at the channel side walls, called the wall effect. The wall effect generates unexpected particle trajectories that may compromise the separation efficiency. In order to anticipate the wall effect when designing DLD devices, a predictive model is proposed in this work and has been validated experimentally. In addition to the usual geometrical parameters, a new parameter, the number of pillars in the channel cross dimension, is considered in this model to investigate its influence on the particle trajectories.     

Object Recognition in X-ray Testing Using Adaptive Sparse Representations

In recent years, X-ray screening systems have been used to safeguard environments in which access control is of paramount importance. Security checkpoints have been placed at the entrances to many public places to detect prohibited items such as handguns and explosives. Human operators complete these tasks because automated recognition in baggage inspection is far from perfect. Research and development on X-ray testing is, however, ongoing into new approaches that can be used to aid human operators. This paper attempts to make a contribution to the field of object recognition by proposing a new approach called Adaptive Sparse Representation (XASR+). It consists of two stages: learning and testing. In the learning stage, for each object of training dataset, several patches are extracted from its X-ray images in order to construct representative dictionaries. A stop-list is used to remove very common words of the dictionaries. In the testing stage, test patches of the test image are extracted, and for each test patch a dictionary is built concatenating the ‘best’ representative dictionary of each object. Using this adapted dictionary, each test patch is classified following the Sparse Representation Classification methodology. Finally, the test image is classified by patch voting. Thus, our approach is able to deal with less constrained conditions including some contrast variability, pose, intra-class variability, size of the image and focal distance. We tested the effectiveness of our method for the detection of four different objects. In our experiments, the recognition rate was more than 97 % in each class, and more than 94 % if the object is occluded less than 15 %. Results show that XASR+ deals well with unconstrained conditions, outperforming various representative methods in the literature.     

An evolutionary algorithm for global optimization based on self-organizing maps

In this article, a new population-based algorithm for real-parameter global optimization is presented, which is denoted as self-organizing centroids optimization (SOC-opt). The proposed method uses a stochastic approach which is based on the sequential learning paradigm for self-organizing maps (SOMs). A modified version of the SOM is proposed where each cell contains an individual, which performs a search for a locally optimal solution and it is affected by the search for a global optimum. The movement of the individuals in the search space is based on a discrete-time dynamic filter, and various choices of this filter are possible to obtain different dynamics of the centroids. In this way, a general framework is defined where well-known algorithms represent a particular case. The proposed algorithm is validated through a set of problems, which include non-separable problems, and compared with state-of-the-art algorithms for global optimization.     

Development of a water-in-oil-in-water multiple emulsion system integrating biomimetic aqueous-core lipid nanodroplets for protein entity stabilization. Part II: process and product characterization

The aqueous-core enclosed in lipid nanoballoons integrating multiple emulsions of the type water-in-oil-in-water mimic, at least in theory, the environment within viable cells, thus being suitable for housing hydrophilic protein entities such as bioactive proteins, peptides and bacteriophage particles. This study reports a complete physicochemical characterization of optimized biomimetic aqueous-core lipid nanoballoons housing hydrophilic (BSA) protein entities, evolved from a statistical 2³×3¹ factorial design study (three variables at two levels and one variable at three levels) that was the subject of the first paper of a series of three, aiming at complete stabilization of the three-dimensional structure of protein entities attempted via housing the said molecular entities within biomimetic aqueous-core lipid nanoballoons integrating a multiple (W/O/W) emulsion. The statistical factorial design followed led to the production of an optimum W/O/W multiple emulsion possessing quite homogeneous particles with an average hydrodynamic size of (186.2?±?2.6) nm and average Zeta potential of (?36.5?±?0.9) mV, and exhibiting a polydispersity index of 0.206?±?0.014. Additionally, the results obtained for the diffusion coefficient of the lipid nanoballoons integrating the optimized W/O/W multiple emulsion were comparable and of the same order of magnitude (10^?12 m² s^?1) as those published by other authors since, typically, diffusion coefficients for molecules range from 10^?10 to 10^?7 m² s^?1, but diffusion coefficients for nanoparticles are typically of the order of magnitude of 10^?12 m² s^?1.     

High-resolution venography of the brain using magnetic resonance imaging

The purpose of this study was to evaluate a non-flow related magnetic resonance imaging method to visualize small veins independent of arteries in the human brain. A long TE, high-resolution 3D gradient echo MR acquisition was used to highlight venous information. The method is based on the paramagnetic property of deoxyhemoglobin and the resulting phase difference between veins and brain parenchyma at long echo times. The MR magnitude images were masked with a phase mask filter to enhance small structure visibility.. Venous information down to sub-pixel vessel diameters of several hundred microns is visible. Venous data are displayed in an angiographic manner using a minimum intensity projection algorithm. Both superficial veins and deep white matter veins are visible. The method has been successfully applied in volunteers. Preliminary results in patients with cerebral arteriovenous malformations indicate its potential in clinical applications. The proposed method is easy to implement and does not require administration of a contrast agent or application of specially designed rf pulses to highlight the veins. Rather it exploits the intrinsic magnetic properties (BOLD-effect) and the prolonged T ^_2* of venous blood. The method may be of diagnostic potential in the assessment of arteriovenous malformations or other vascular venous lesions. © 1998 Elsevier Science B.V. All rights reserved.