Xiang Zhang

Bio: Xiang Zhang is an academic researcher from Baylor College of Medicine. The author has contributed to research in topics: Medicine & Computer science. The author has an hindex of 154, co-authored 1733 publications receiving 117576 citations. Previous affiliations of Xiang Zhang include University of California, Berkeley & University of Texas MD Anderson Cancer Center.

Constraints on the off-shell Higgs boson signal strength in the high-mass ZZ and WW final states with the ATLAS detector

Abstract: Measurements of the ZZ and WW final states in the mass range above the [Formula: see text] and [Formula: see text] thresholds provide a unique opportunity to measure the off-shell coupling strength of the Higgs boson. This paper presents constraints on the off-shell Higgs boson event yields normalised to the Standard Model prediction (signal strength) in the [Formula: see text], [Formula: see text] and [Formula: see text] final states. The result is based on pp collision data collected by the ATLAS experiment at the LHC, corresponding to an integrated luminosity of 20.3 fb[Formula: see text] at a collision energy of [Formula: see text] TeV. Using the [Formula: see text] method, the observed 95 [Formula: see text] confidence level (CL) upper limit on the off-shell signal strength is in the range 5.1-8.6, with an expected range of 6.7-11.0. In each case the range is determined by varying the unknown [Formula: see text] and [Formula: see text] background K-factor from higher-order quantum chromodynamics corrections between half and twice the value of the known signal K-factor. Assuming the relevant Higgs boson couplings are independent of the energy scale of the Higgs boson production, a combination with the on-shell measurements yields an observed (expected) 95 [Formula: see text] CL upper limit on [Formula: see text] in the range 4.5-7.5 (6.5-11.2) using the same variations of the background K-factor. Assuming that the unknown [Formula: see text] background K-factor is equal to the signal K-factor, this translates into an observed (expected) 95 [Formula: see text] CL upper limit on the Higgs boson total width of 22.7 (33.0) MeV.

Exciton-dominant electroluminescence from a diode of monolayer MoS2

Abstract: In two-dimensional monolayer MoS2, excitons dominate the absorption and emission properties. However, the low electroluminescent efficiency and signal-to-noise ratio limit our understanding of the excitonic behavior of electroluminescence. Here, we study the microscopic origin of the electroluminescence from a diode of monolayer MoS2 fabricated on a heavily p-type doped silicon substrate. Direct and bound-exciton related recombination processes are identified from the electroluminescence. At a high electron-hole pair injection rate, Auger recombination of the exciton-exciton annihilation of the bound exciton emission is observed at room temperature. Moreover, the efficient electrical injection demonstrated here allows for the observation of a higher energy exciton peak of 2.255 eV in the monolayer MoS2 diode, attributed to the excited exciton state of a direct-exciton transition.

Prediction of stiffness and stresses in z-fibre reinforced composite laminates

Abstract: The mechanical properties of z-pinned composite laminates were examined numerically. Finite element calculations have been performed to understand how the through-thickness reinforcement modifies the engineering elastic constants and local stress distributions. Solutions were found for four basic laminate stacking sequences, all having two percent volume fraction of z-fibres. For the stiffness analysis, a micro-mechanical finite element model was employed that was based on the actual geometric configuration of a z-pinned composite unit cell. The numerical results agreed very well with some published solutions. It showed that by adding 2% volume fraction of z-fibres, the through-thickness Young's modulus was increased by 22–35%. The reductions in the in-plane moduli were contained within 7–10%. The stress analysis showed that interlaminar stress distributions near a laminate free edge were significantly affected when z-fibres were placed within a characteristic distance of one z-fibre diameter from the free edge. Local z-fibres carried significant amount of interlaminar normal and shear stresses.

Control of Coherently Coupled Exciton Polaritons in Monolayer Tungsten Disulphide.

Abstract: Monolayer transition metal dichalcogenides (TMD) with confined 2D Wannier-Mott excitons are intriguing for the fundamental study of strong light-matter interactions and the exploration of exciton polaritons at high temperatures. However, the research of 2D exciton polaritons has been hindered because the polaritons in these atomically thin semiconductors discovered so far can hardly support strong nonlinear interactions and quantum coherence due to uncontrollable polariton dynamics and weakened coherent coupling. In this work, we demonstrate, for the first time, a precisely controlled hybrid composition with angular dependence and dispersion-correlated polariton emission by tuning the polariton dispersion in TMD over a broad temperature range of 110--230 K in a single cavity. This tamed polariton emission is achieved by the realization of robust coherent exciton-photon coupling in monolayer tungsten disulphide (${\mathrm{WS}}_{2}$) with large splitting-to-linewidth ratios ($g3.3$). The unprecedented ability to manipulate the dispersion and correlated properties of TMD exciton polaritons at will offers new possibilities to explore important quantum phenomena such as inversionless lasing, Bose-Einstein condensation, and superfluidity.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Going deeper with convolutions

Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.