Search

In the operation of OFETs, the electrical properties are strongly dependent on the merits of the constituting layers and the formed interfaces. Here we study the trap states variations at the interface between the organic semiconductor pentacene and polymer insulators. With ZrO2 dielectric modified by polymers and find a 10 × decrease in the density of trap states at the semiconductor/insulator interface, bring about the charge carrier mobility increase from 0.058 cm2/Vs to 0.335 cm2/Vs. In addition, when compare to the thicker films at the same applied gate voltage, the thinner film would lead to enhanced coupling capability and more charges cumulative cumulated at the channel region, which is pivotal for optimizing the performance of OFETs. The results prove that the property of the insulator layer could impact largely on the device performance.

Small target detection is a crucial technique that restricts the performance of many infrared imaging systems. In this paper, a novel detection model of infrared small target via non-convex tensor rank surrogate joint local contrast energy (NTRS) is proposed. To improve the latest infrared patch-tensor (IPT) model, a non-convex tensor rank surrogate merging tensor nuclear norm (TNN) and the Laplace function, is utilized for low rank background patch-tensor constraint, which has a useful property of adaptively allocating weight for every singular value and can better approximate l 0 -norm. Considering that the local prior map can be equivalent to the saliency map, we introduce a local contrast energy feature into IPT detection framework to weight target tensor, which can efficiently suppress the background and preserve the target simultaneously. Besides, to remove the structured edges more thoroughly, we suggest an additional structured sparse regularization term using the l 1 , 1 , 2 -norm of third-order tensor. To solve the proposed model, a high-efficiency optimization way based on alternating direction method of multipliers with the fast computing of tensor singular value decomposition is designed. Finally, an adaptive threshold is utilized to extract real targets of the reconstructed target image. A series of experimental results show that the proposed method has robust detection performance and outperforms the other advanced methods.

Infrared (IR) small target detection in complex background is one of the key techniques in many defense systems. In this paper, a novel method based on infrared patch-tensor model with structured sparse regularization (IPTSS) is proposed for small IR target detection. To overcome the structured edge interferences , the IPTSS model adds an edge structured sparse item utilizing the e1,1,2 norm minimization constraint to the IPT model based on partial sum of tensor nuclear norm (PSTNN). An efficient optimization algorithm based on alternating direction method of multipliers (ADMM) is designed to solve the proposed IPTSS model. After the target-background-edge components are separated, the target image can be reconstructed. Finally, small targets can be extracted easily via adaptive threshold segmentation in the reconstructed target image. Extensive experimental results demonstrate that the proposed method can effectively enhance small IR targets while suppressing the complex background in various scenes.

To improve the detection ability of infrared small targets in complex backgrounds, a novel method based on non-convex rank approximation minimization joint l2,1 norm (NRAM) was proposed. Due to the defects of the nuclear norm and l1 norm, the state-of-the-art infrared image-patch (IPI) model usually leaves background residuals in the target image. To fix this problem, a non-convex, tighter rank surrogate and weighted l1 norm are instead utilized, which can suppress the background better while preserving the target efficiently. Considering that many state-of-the-art methods are still unable to fully suppress sparse strong edges, the structured l2,1 norm was introduced to wipe out the strong residuals. Furthermore, with the help of exploiting the structured norm and tighter rank surrogate, the proposed model was more robust when facing various complex or blurry scenes. To solve this non-convex model, an efficient optimization algorithm based on alternating direction method of multipliers (ADMM) plus difference of convex (DC) programming was designed. Extensive experimental results illustrate that the proposed method not only shows superiority in background suppression and target enhancement, but also reduces the computational complexity compared with other baselines.

Background Studies which have attempted to assess the predictive potential of socio-environmental risk factors for psychosis have used such a variety of datasets and methodologies. As a result, it is not possible for policy-makers to understand how different models compare, or might inform evidence-based policy-making. Thus, the cumulative predictive potential of non-genetic risk factors for psychosis has not yet been studied systematically. An important question which has not been considered previously is whether correlation structure between multivariate risks can be detrimental to the goal of prediction, particularly across different populations. Model fitting to locally-relevant correlation structures can limit the generalizability of a prediction model. Copulas are mathematical functions which allow the joint risk function of two or more correlated variables to be modelled in spite of this inherent bias. The copula approach is a foundation methodology with applications in the fields of finance, insurance and banking, where it is used for risk-management purposes. This study examines the impact of copulas on the stability of prediction power for psychosis across different populations. Methods The data used in this work comes from work package 2 (WP2) (entitled “Functional Enviromics”) of The European Union (EU)-funded European Network Study of Gene-Environment Interactions (EUGEI). The total dataset available consists of 1180 cases of first episode psychosis (ICD10 diagnostic criteria F20-F29 or F30-F33) and 1528 healthy controls recruited by 16 centres across 6 countries (United Kingdom, Holland, Spain, France, Italy, Brazil). We sought to compare the predictive performance of copulas against that of summary risk scores for formulating disease risk for a common set of socio-environmental risk factors. The copula methodology allows joint risks to be modelled as a distribution whilst summary scores convey the number of risk factors encountered by an individual, weighted by literature-derived odds ratios for association. Gaussian copula with non-Gaussian marginal distributions were used to capture the correlation structure of 9 discrete variables in total. These incorporated: Lifetime Cannabis Use, frequency of Cannabis Use, Household discord, severity of psychological abuse, severity of physical abuse, severity of sexual abuse, severity of bullying, number of adverse adult life events and intrusive adult life events. We applied a fully Bayesian approach which uses Markov Chain Monte Carlo to simulate latent variables from multivariate ordered probit model and also estimate the threshold parameters and parameters from copula model. The resulting joint distribution (a copula) mapped the relationship between cumulative exposure to these factors and risk of psychosis. Results A proportion of subjects were withheld from the copula, so that the performance of the finished function could be evaluated on unseen data. The performance of the 2 prediction methods was compared within and between recruitment centres and are conveyed in terms of: • Sensitivity and false positive rates (The area under the Receiver Operating Characteristic curve) • Percentage of variance explained (Nagelkerke R2) • Calibration (whether predicted risks were correct) • Discrimination (whether high risk subjects could be distinguished from low risk ones) • Reclassification (model behaviour close to specific thresholds) Discussion The application of the copula methodology to the multi-centre EUGEI dataset provides us with the opportunity to tackle a major limitation of the summary scoring approach which is the default method for aggregating risks across most areas of health research.