University of Texas at Dallas

About: University of Texas at Dallas is a education organization based out in Richardson, Texas, United States. It is known for research contribution in the topics: Population & Computer science. The organization has 14986 authors who have published 35589 publications receiving 1293714 citations. The organization is also known as: UT-Dallas & UT Dallas.

High-efficiency carrier multiplication through direct photogeneration of multi-excitons via virtual single-exciton states

Abstract: We have previously demonstrated that absorption of a single photon by a nanocrystal quantum dot can generate multiple excitons with an efficiency of up to 100%. This effect, known as carrier multiplication, should lead to substantial improvements in the performance of a variety of optoelectronic and photocatalytic devices, including solar cells, low-threshold lasers and entangled photon sources. Here we present detailed analysis of the dynamics that govern the ultrafast growth of multi-exciton populations in CdSe and PbSe nanocrystals and propose a model of how such populations arise. Our analysis indicates that the generation of multi-excitons in these systems takes less than 200 fs, which suggests that it is an instantaneous event. We explain this in terms of their direct photogeneration via multiple virtual single-exciton states. This process relies on both the confinement-enhanced Coulomb coupling between single excitons and multi-excitons and the large spectral density of high-energy single- and multi-exciton resonances that occur in semiconductor nanocrystals.

DeepRoad: GAN-based metamorphic testing and input validation framework for autonomous driving systems

Abstract: While Deep Neural Networks (DNNs) have established the fundamentals of image-based autonomous driving systems, they may exhibit erroneous behaviors and cause fatal accidents. To address the safety issues in autonomous driving systems, a recent set of testing techniques have been designed to automatically generate artificial driving scenes to enrich test suite, e.g., generating new input images transformed from the original ones. However, these techniques are insufficient due to two limitations: first, many such synthetic images often lack diversity of driving scenes, and hence compromise the resulting efficacy and reliability. Second, for machine-learning-based systems, a mismatch between training and application domain can dramatically degrade system accuracy, such that it is necessary to validate inputs for improving system robustness. In this paper, we propose DeepRoad, an unsupervised DNN-based framework for automatically testing the consistency of DNN-based autonomous driving systems and online validation. First, DeepRoad automatically synthesizes large amounts of diverse driving scenes without using image transformation rules (e.g. scale, shear and rotation). In particular, DeepRoad is able to produce driving scenes with various weather conditions (including those with rather extreme conditions) by applying Generative Adversarial Networks (GANs) along with the corresponding real-world weather scenes. Second, DeepRoad utilizes metamorphic testing techniques to check the consistency of such systems using synthetic images. Third, DeepRoad validates input images for DNN-based systems by measuring the distance of the input and training images using their VGGNet features. We implement DeepRoad to test three well-recognized DNN-based autonomous driving systems in Udacity self-driving car challenge. The experimental results demonstrate that DeepRoad can detect thousands of inconsistent behaviors for these systems, and effectively validate input images to potentially enhance the system robustness as well.

Observation of Associated Near-Side and Away-Side Long-Range Correlations in √sNN=5.02 TeV Proton-Lead Collisions with the ATLAS Detector

Abstract: Two-particle correlations in relative azimuthal angle (Delta phi) and pseudorapidity (Delta eta) are measured in root S-NN = 5.02 TeV p + Pb collisions using the ATLAS detector at the LHC. The measurements are performed using approximately 1 mu b(-1) of data as a function of transverse momentum (p(T)) and the transverse energy (Sigma E-T(Pb)) summed over 3.1 < eta < 4.9 in the direction of the Pb beam. The correlation function, constructed from charged particles, exhibits a long-range (2 < vertical bar Delta eta vertical bar < 5) "near-side" (Delta phi similar to 0) correlation that grows rapidly with increasing Sigma E-T(Pb). A long-range "away-side" (Delta phi similar to pi) correlation, obtained by subtracting the expected contributions from recoiling dijets and other sources estimated using events with small Sigma E-T(Pb), is found to match the near-side correlation in magnitude, shape (in Delta eta and Delta phi) and Sigma E-T(Pb) dependence. The resultant Delta phi correlation is approximately symmetric about pi/2, and is consistent with a dominant cos2 Delta phi modulation for all Sigma E-T(Pb) ranges and particle p(T).

The Performance Consequences of Ambidexterity in Strategic Alliance Formations: Empirical Investigation and Computational Theorizing

Abstract: Although alliance studies have generally favored an ambidextrous approach between exploration and exploitation, they tend to overlook a firm's characteristics, its industry constraints, or the dynamic network in which the firm is embedded. This study examines the ambidexterity hypothesis and its boundary conditions with a unique research method. We not only analyze empirical data from five U.S. industries spanning eight years, but also expand theoretical insights to the network level by building a computer simulation model. Both our empirical and simulation results reveal the contingencies of the ambidexterity hypothesis in alliance formation. Our findings show that although an ambidextrous formation of alliances benefits large firms, a focused formation of either exploratory or exploitative alliances benefits small firms. In an uncertain environment an ambidextrous formation enhances firm performance but so does a focused formation in a stable environment. Finally, the simulation model demonstrates that a firm's centrality and structural hole positions in network relations can moderate the relationships between alliance formation choices and firm performance, and that the ambidexterity hypothesis may be limited to the earlier stage of the network. Our study provides critical evidence into the viability of adopting a dynamic network perspective in understanding the ambidexterity hypothesis and advancing strategic alliance research beyond static and dyadic levels.