Engagnition: A multi-dimensional dataset for engagement recognition of children with autism spectrum disorder.

Engagement plays a key role in improving the cognitive and motor development of children with autism spectrum disorder (ASD). Sensing and recognizing their engagement is crucial before sustaining and improving the engagement. Engaging technologies involving interactive and multi-sensory stimuli have improved engagement and alleviated hyperactive and stereotyped behaviors. However, due to the scarcity of data on engagement recognition for children with ASD, limited access to and small pools of participants, and the prohibitive application requirements such as robots, high cost, and expertise, implementation in real world is challenging. However, serious games have the potential to overcome those drawbacks and are suitable for practical use in the field. This study proposes Engagnition, a dataset for engagement recognition of children with ASD (N = 57) using a serious game, "Defeat the Monster," based on enhancing recognition and classification skills. The dataset consists of physiological and behavioral responses, annotated by experts. For technical validation, we report the distributions of engagement and intervention, and the signal-to-noise ratio of physiological signals.

A Machine Learning-Based Initial Difficulty Level Adjustment Method for Balance Exercise on a Trunk Rehabilitation Robot.

Trunk rehabilitation exercises such as those for remediating core stability can help improve the seated balance of patients with weakness or loss of proprioception caused by diseases such as stroke, and aid the recovery of other functions such as gait. However, there has not yet been any reported method for automatically determining the parameters that define exercise difficulty on a trunk rehabilitation robot (TRR) based on data such as the patient's demographic information, balancing ability, and training sequence, etc. We have proposed a machine learning (ML)-based difficulty adjustment method to determine an appropriate virtual damping gain (Dvirtual) of the controller for the TRR's unstable training mode. Training data for the proposed system is obtained from 37 healthy young adults, and the trained ML model thus obtained is tested through experiments with a separate population of 25 healthy young adults. The leave-one-out cross validation results (37 subjects) from the training group for validation of the designed ML model showed 80.90% average accuracy (R2 score) for using the given information to predict the desired difficulty levels, which are represented by the level of balance performance quantified as Mean Velocity Displacement (MVD) of the center of pressure. Statistical analysis (Repeated measures analysis of variance) of subject performance also showed that ground truth difficulty levels from the training data and predicted difficulty levels did not differ significantly under any of the three exercise modes used in this study (Hard, Medium, and Easy), and the standard deviations were reduced by 16.39, 41.39, and 25.68%, respectively. Moreover, the Planar Deviation (PD) of the center of pressure, which was not the target parameter here, also showed results similar to the MVD, which indicates that the predicted Dvirtual affected the difficulty level of balance performance. Therefore, the proposed ML model-based difficulty adjustment method has potential for use with people who have varied balancing abilities.

Highly Efficient and Stable Iridium Oxygen Evolution Reaction Electrocatalysts Based on Porous Nickel Nanotube Template Enabling Tandem Devices with Solar-to-Hydrogen Conversion Efficiency Exceeding 10.

Ir is one of the most efficient oxygen evolution reaction (OER) catalysts; however, it is also one of the rarest and most expensive elements. Therefore, it is highly desirable to develop Ir catalysts with nanostructures that reduce Ir consumption by maximizing the surface-to-volume ratio without limiting the mass transport of reactants and products of reactions. Ir OER catalysts on a template that consisted of porous nanotubes (PNTs) based on Ni are fabricated. The Ir/Ni PNTs offer multiple benefits, including high catalytic performance (potential of 1.500 V vs. reversible hydrogen electrode (RHE) at an operating current density of 10 mA cm-2 and Tafel slope of 44.34 mV decade-1 ), minimal use of Ir (mass activity of 3273 A g-1 at 1.53 V vs RHE), and facile mass transport through the NT-sidewall pores (stable operation for more than 10 h). The Ir/Ni PNTs are also applied to a tandem device, consisting of a Cu(In,Ga)Se2 -based photocathode and halide perovskite photovoltaic cell, for unassisted water splitting. A solar-to-hydrogen conversion efficiency that exceeded 10% is also demonstrated, which is nearly 1% point greater than when a planar Ir film is used as the anode instead of Ir/Ni PNTs.

Current Status of Amebic Liver Abscess in Korea Comparing with Pyogenic Liver Abscess.

Backgrounds/Aims: With the improvement of hygiene, the incidence of amebic liver abscess is decreasing in South Korea. On the other hand, there is little data on the status of amebic liver abscess compared to pyogenic liver abscess. Methods: Patients with an amebic liver abscess, in whom Entamoeba histolytica (E. histolytica) IgG was positive, were identified retrospectively in a university hospital. The clinical, laboratory, and radiological characteristics of amebic liver abscess were compared with those of pyogenic liver abscess in the same period. Results: Between March 2010 and October 2016, 413 patients with a liver abscess were identified. Among them, the serologic test for E. histolytica was performed in 209 patients. Fifteen (7.2%) were classified as an amebic liver abscess, and the remainder were diagnosed with a pyogenic liver abscess. The age, gender, white blood cell, and CRP was comparable between the two groups. Procalcitonin was lower in amebic liver abscess than the pyogenic one. On CT, peripheral rim enhancement was more frequent, but cluster signs were not observed in amebic liver abscess compared to pyogenic liver abscess. None of the patients with amebic liver abscess died. In contrast, the mortality of pyogenic liver abscess was 4.7%. Conclusions: Amebic liver abscess should still be considered as one of the causes of liver abscess in Korea. It is difficult to discriminate an amebic liver abscess from a pyogenic liver abscess only according to the clinical, laboratory, and radiologic findings. Therefore, it is necessary to perform a serologic test for E. histolytica for a precise evaluation of liver abscess in a high-risk group.

Improvement of power conversion efficiency by a stepwise band-gap structure for silicon quantum dot solar cells.

As a promising next-generation solar cell, the power conversion efficiency of a silicon quantum dot (Si-QD) solar cell is still low. In this work, the band-gap structure of a Si-QD layer was modified to improve the power conversion efficiency of a Si-QD solar cell. A stepwise band-gap Si-QD (SB Si-QD) layer with a high bandgap top layer (about 2.22 eV) and a low band-gap bottom layer (about 1.98 eV) was grown on a Si (100) substrate. The open circuit voltage and short circuit current were improved by band-gap engineering of the Si-QD absorption layer. As a result, the power conversion efficiency of the SB Si-QD solar cell increased from 16.50% to 17.50%, compared to that of a Si-QD solar cell with a uniform band gap. This results will provide a guide to design advanced Si-QD solar cells by considering the band-gap structure in the Si-QD absorption layer.

High efficiency Si quantum dot heterojunction solar cells using a single SiOX:B layer.

Si quantum dots (QDs) have been fabricated from SiO2/SiOx multilayer structures to create a homogeneous size. However, this structure achieved much lower efficiencies than would be expected in the Si QD photovoltaic field. This is because Si QD generation and photoexcited carrier transport is restricted by the adjacent SiO2 layer. In this study, we applied a single SiOx:B layer fabrication method to the Si QD heterojunction solar cells. The number of generated Si QDs and the photo-excited carrier lifetime was maximized when the oxygen partial pressure and boron doping concentration parameters were 2.7 × 10-5 Torr and 2.27 × 1021 atoms cm-3, respectively. As a result, in excess of 17% power conversion efficiency for the Si QD heterojunction solar cell was achieved using the single layer method.

Ultraviolet responses of a heterojunction Si quantum dot solar cell.

We investigated the ultraviolet (UV) responses of a heterojunction Si quantum dot (QD) solar cell consisting of p-type Si-QDs fabricated on a n-type crystalline Si (p-Si-QD/n-c-Si HJSC). The UV responses were compared with a conventional n-type crystalline Si solar cell (n-c-Si SC). The external and internal quantum efficiency results of the p-Si-QD/n-c-Si HJSC exhibited a clear enhancement in the UV responses (300-400 nm), which was not observed in the n-c-Si SC. Based on the results of the cell reflectance and bias-dependent responses, we expect that almost all UV responses occur in the p-Si-QD layer, and the generated carriers can be transported via the Si-QD layer due to the formation of a sufficient electric filed. As a result, a high power conversion efficiency of 14.5% was achieved from the p-Si-QD/n-c-Si HJSC. By reducing the thickness of the n-Si substrate from 650 µm to 300 µm, more enhanced power conversion efficiency of 14.8% was obtained which is the highest value among the reported Si-QD based solar cells to date.

Bacterial analysis by laser desorption ionization mass spectrometry on amorphous silicon.

Lipid profiling in nine bacterial species has been accomplished by laser desorption ionization mass spectrometry (LDI-MS) using amorphous silicon (a-Si) thin film with 100 nm thickness. Lipid ions could be generated by LDI on a-Si regardless of ion acquisition modes because of a thermal property of a-Si to govern laser-induced surface heating. In a comparative study of lipid profiling in Bacillus lichemiformis by LDI-MS and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS), LDI-MS on a-Si shows a higher efficiency in lipid and lipopeptide detection than MALDI-MS. A total of 53 peaks of lipid ions generated by LDI on a-Si in both acquisition modes for m/z 400-1200 was 1.6 times more than that detected by MALDI-MS using three organic matrices-2,5-dihydroxybenzoic acid, 1,5-diaminonaphthalene, and 2,4,6-trihydroxyacetophenone monohydrate. Also, the authors demonstrate by mass spectrometry imaging (MSI) that LDI-MS provides high detection coverage through whole sample area. MSI results show the detection yield in LDI on a-Si is 94.8% calculated by counting the number of points detected in the analyte ion signal in a whole spot. It means that reproducible detection of lipid ions by LDI-MS is possible even if laser is randomly irradiated at any position within the bacterial sample area applied on a-Si. Lipid profiling by LDI-MS on a-Si was applied to bacterial differentiation of nine bacterial species conducted by performing principal component analysis. Nine bacterial species are successfully distinguishable from each other by LDI-MS lipid profiling.

Active Player Modeling in the Iterated Prisoner's Dilemma.

The iterated prisoner's dilemma (IPD) is well known within the domain of game theory. Although it is relatively simple, it can also elucidate important problems related to cooperation and trust. Generally, players can predict their opponents' actions when they are able to build a precise model of their behavior based on their game playing experience. However, it is difficult to make such predictions based on a limited number of games. The creation of a precise model requires the use of not only an appropriate learning algorithm and framework but also a good dataset. Active learning approaches have recently been introduced to machine learning communities. The approach can usually produce informative datasets with relatively little effort. Therefore, we have proposed an active modeling technique to predict the behavior of IPD players. The proposed method can model the opponent player's behavior while taking advantage of interactive game environments. This experiment used twelve representative types of players as opponents, and an observer used an active modeling algorithm to model these opponents. This observer actively collected data and modeled the opponent's behavior online. Most of our data showed that the observer was able to build, through direct actions, a more accurate model of an opponent's behavior than when the data were collected through random actions.

Recent Advances in General Game Playing.

The goal of General Game Playing (GGP) has been to develop computer programs that can perform well across various game types. It is natural for human game players to transfer knowledge from games they already know how to play to other similar games. GGP research attempts to design systems that work well across different game types, including unknown new games. In this review, we present a survey of recent advances (2011 to 2014) in GGP for both traditional games and video games. It is notable that research on GGP has been expanding into modern video games. Monte-Carlo Tree Search and its enhancements have been the most influential techniques in GGP for both research domains. Additionally, international competitions have become important events that promote and increase GGP research. Recently, a video GGP competition was launched. In this survey, we review recent progress in the most challenging research areas of Artificial Intelligence (AI) related to universal game playing.

Graphene/Si-quantum-dot heterojunction diodes showing high photosensitivity compatible with quantum confinement effect.

Graphene/Si quantum dot (QD) heterojunction diodes are reported for the first time. The photoresponse, very sensitive to variations in the size of the QDs as well as in the doping concentration of graphene and consistent with the quantum-confinement effect, is remarkably enhanced in the near-ultraviolet range compared to commercially available bulk-Si photodetectors. The photoresponse proves to be dominated by the carriertunneling mechanism.

Inference of other's internal neural models from active observation.

Recently, there have been several attempts to replicate theory of mind, which explains how humans infer the mental states of other people using multiple sensory input, with artificial systems. One example of this is a robot that observes the behavior of other artificial systems and infers their internal models, mapping sensory inputs to the actuator's control signals. In this paper, we present the internal model as an artificial neural network, similar to biological systems. During inference, an observer can use an active incremental learning algorithm to guess an actor's internal neural model. This could significantly reduce the effort needed to guess other people's internal models. We apply an algorithm to the actor-observer robot scenarios with/without prior knowledge of the internal models. To validate our approach, we use a physics-based simulator with virtual robots. A series of experiments reveal that the observer robot can construct an "other's self-model", validating the possibility that a neural-based approach can be used as a platform for learning cognitive functions.

Amplitude death of coupled hair bundles with stochastic channel noise.

Hair cells conduct auditory transduction in vertebrates. In lower vertebrates such as frogs and turtles, due to the active mechanism in hair cells, hair bundles (stereocilia) can be spontaneously oscillating or quiescent. Recently an amplitude death phenomenon has been proposed [K.-H. Ahn, J. R. Soc. Interface, 10, 20130525 (2013)] as a mechanism for auditory transduction in frog hair-cell bundles, where sudden cessation of the oscillations arises due to the coupling between nonidentical hair bundles. The gating of the ion channel is intrinsically stochastic due to the stochastic nature of the configuration change of the channel. The strength of the noise due to the channel gating can be comparable to the thermal Brownian noise of hair bundles. Thus, we perform stochastic simulations of the elastically coupled hair bundles. In spite of stray noisy fluctuations due to its stochastic dynamics, our simulation shows the transition from collective oscillation to amplitude death as interbundle coupling strength increases. In its stochastic dynamics, the formation of the amplitude death state of coupled hair bundles can be seen as a sudden suppression of the displacement fluctuation of the hair bundles as the coupling strength increases. The enhancement of the signal-to-noise ratio through the amplitude death phenomenon is clearly seen in the stochastic dynamics. Our numerical results demonstrate that the multiple number of transduction channels per hair bundle is an important factor to the amplitude death phenomenon, because the phenomenon may disappear for a small number of transduction channels due to strong gating noise.

In situ control of oxygen vacancies in TiO2 by atomic layer deposition for resistive switching devices.

Oxygen vacancies (V(O)) have profound effects on the physical and chemical performance of devices based on oxide materials. This is particularly true in the case of oxide-based resistive random access memories, in which memory switching operation under an external electrical stimulus is closely associated with the migration and ordering of the oxygen vacancies in the oxide material. In this paper, we report on a reliable approach to in situ control of the oxygen vacancies in TiOx films. Our strategy for tight control of the oxygen vacancy is based on the utilization of plasma-enhanced atomic layer deposition of titanium oxide under precisely regulated decomposition of the precursor molecules (titanium (IV) tetraisopropoxide, Ti[OCH(CH3)2]4) by plasma-activated reactant mixture (N2+O2). From the various spectroscopic and microstructural analyses by using Rutherford backscattering spectrometry, x-ray photoelectron spectroscopy, high-resolution transmission electron microscopy, confocal Raman spectroscopy, and spectroscopic ellipsometry, we found that the precursor decomposition power (R(F)) of plasma-activated reactant mixture determines not only the oxygen vacancy concentration but also the crystallinity of the resulting TiO(x) film: nanocrystalline anatase TiO(x) with fewer oxygen vacancies under high R(F), while amorphous TiOx with more oxygen vacancies under low RF. Enabled by our controlling capability over the oxygen vacancy concentration, we were able to thoroughly elucidate the effect of oxygen vacancies on the resistive switching behavior of TiO(x)-based memory capacitors (Pt/TiO(x)/Pt). The electrical conduction behavior at the high resistance state could be explained within the framework of the trap-controlled space-charge-limited conduction with two characteristic transition voltages. One is the voltage (V(SCL)) for the transition from Ohmic conduction to space-charge-limited conduction, and the other is the voltage (V(TFL)) for transition from space-charge-limited conduction to trap-filled-limited conduction. In this work, we have disclosed for the first time the dependence of these two characteristic transition voltages (i.e., V(SCL) and V(TFL)) on the oxygen vacancy concentration.

In situ determination of the pore opening point during wet-chemical etching of the barrier layer of porous anodic aluminum oxide: nonuniform impurity distribution in anodic oxide.

Wet-chemical etching of the barrier oxide layer of anodic aluminum oxide (AAO) was systematically investigated by using scanning electron microscopy (SEM), secondary ion mass spectrometry (SIMS), and a newly devised experimental setup that allows accurate in situ determination of the pore opening point during chemical etching of the barrier oxide layer. We found that opening of the barrier oxide layer by wet-chemical etching can be significantly influenced by anodization time (tanodi). According to secondary ion mass spectrometry (SIMS) analysis, porous anodic aluminum oxide (AAO) samples formed by long-term anodization contained a lower level of anionic impurity in the barrier oxide layer compared to the short-term anodized one and consequently exhibited retarded opening of the barrier oxide layer during the wet-chemical etching. The observed compositional dependence on the anodization time (tanodi) in the barrier oxide layer is attributed to the progressive decrease of the electrolyte concentration upon anodization. The etching rate of the outer pore wall at the bottom part is lower than that of the one at the top part due to the lower level of impurity content in that region. This indicates that a concentration gradient of anionic impurity in the outer pore wall oxide may be established along both the vertical and radial directions of cylindrical pores. Apart from the effect of electrolyte concentration on the chemical composition of the barrier oxide layer, significantly decreased current density arising from the lowered concentration of electrolyte during the long-term anodization (~120 h) was found to cause disordering of pores. The results of the present work are expected to provide viable information not only for practical applications of nanoporous AAO in nanotechnology but also for thorough understanding of the self-organized formation of oxide nanopores during anodization.

Optical and electronic properties of hydrogenated silicon nanoclusters and nitrogen passivated silicon nanoclusters: a density functional theory study.

Silicon nanoclusters have become significant research interest due to their potential application to optoelectronic devices in visible range. We investigate the electronic and optical properties of hydrogenated and nitrogen-passivated silicon nanoclusters using density functional theory calculations. The energy gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of nanoclusters have varying sizes. They are systematically studied using the conventional local density approximation, the generalized gradient approximation, and the time-dependent density functional theory calculations with hybrid functional. The HOMO-LUMO gap is found to decrease monotonically as the size of nanocluster increases. Introducing one and two nitrogen passivants to a Si29H36 nanocluster, we find that the HOMO-LUMO gap decreases as the number of nitrogen passivants increases. It suggests that multi-nitrogen passivants may enable light emission in visible range from smaller clusters.

Active doping of B in silicon nanostructures and development of a Si quantum dot solar cell.

Active doping of B was observed in nanometer silicon layers confined in SiO(2) layers by secondary ion mass spectrometry (SIMS) depth profiling analysis and confirmed by Hall effect measurements. The uniformly distributed boron atoms in the B-doped silicon layers of [SiO(2) (8 nm)/B-doped Si(10 nm)](5) films turned out to be segregated into the Si/SiO(2) interfaces and the Si bulk, forming a distinct bimodal distribution by annealing at high temperature. B atoms in the Si layers were found to preferentially substitute inactive three-fold Si atoms in the grain boundaries and then substitute the four-fold Si atoms to achieve electrically active doping. As a result, active doping of B is initiated at high doping concentrations above 1.1 × 10(20) atoms cm( - 3) and high active doping of 3 × 10(20) atoms cm( - 3) could be achieved. The active doping in ultra-thin Si layers was implemented for silicon quantum dots (QDs) to realize a Si QD solar cell. A high energy-conversion efficiency of 13.4% was realized from a p-type Si QD solar cell with B concentration of 4 × 10(20) atoms cm( - 3).

Size- and doping-dependent time-resolved photoluminescence of doped Si nanocrystals.

Time-resolved photoluminescence (PL) has been studied for B- and Sb-doped Si nanocrystals (NCs) fabricated by ion beam sputtering and annealing. For B-doped Si NCs, the PL intensity as well as the PL lifetime (τPL) increases as NC size (d) varies from 1.5 to 2.6 nm, similar to the case for undoped Si NCs, but with further increase of d, they decrease, possibly resulting from the increase of optically less active NCs with the increase of NCs containing more dopants. The PL intensity and τPL monotonically decrease with increasing doping concentration (nD), irrespective of doping element. Si NCs show smaller τPL in B doping than in Sb doping over the full range of nD. The sharp decrease in PL intensity, accompanied by the gradual decrease in τPL for the higher nD of Sb, may be attributed to Auger recombination due to the presence of Sb inside Si NCs. The higher PL quench rate by Sb compared to B could be attributed to better ionization of Sb dopants in Si NCs.

Nonvolatile floating-gate memories using Zr and ZrO2 nanodots.

Triple-layer structures of SiO2/Zr nanodots (NDs)/SiO2 for nonvolatile memories have been firstly fabricated at room temperature by using ion beam sputtering deposition (IBSD). High-resolution transmission electron microscopy and X-ray photoelectron spectroscopy demonstrate that Zr NDs self-assembled between the SiO2 layers by IBSD are changed into ZrO2 NDs by annealing. The memory window that is estimated by capacitance-voltage curves increases up to a maximum value of 5.8 V with increasing Zr amount up to 6 monolayers for the annealed samples. The memory window and the charge-loss rate at the programmed state are smaller before annealing, which is explained with reference to double oxide barriers of SiO2 and ZrO2.

Control of amorphous silica nanowire growth by oxygen content of Si-rich oxide.

Ni-coated Si-rich oxide (SRO, SiO(x)) on a p-type Si wafer has been annealed with Si powder to grow silica nanowires (NWs), which have a composition of stoichiometric SiO(2), irrespective of x. The diameters of the NWs are well controlled from 82 to 23 nm by increasing x from 0.4 to 1.2 and they have a uniform distribution at a fixed x. These results suggest that the oxygen content (x) plays a crucial role in determining the diameter of the NWs at the early stage of the NW formation. The growth behaviors of the NWs are explained well based on a modified vapor-liquid-solid mechanism.