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.

An experimental study of residual stress and direction-dependence of fatigue crack growth behaviour in as-built and stress-relieved selective-laser-melted Ti6Al4V

Abstract: Selective-laser-melting (SLM) is a powder-bed fusion additive-manufacturing process that has the potential to deliver three-dimensional complex parts with mechanical properties comparable or superior to parts produced via traditional manufacturing using cast and wrought alloys. Concerns for metallic parts built via SLM are the process-induced residual stresses, and anisotropic mechanical properties. This paper investigates the effect of residual stresses on the fatigue crack growth rate of SLM Ti6Al4V in as-built and stress-relieved conditions. Neutron diffraction and the contour method are employed to measure residual stresses in compact-tension samples. Neutron diffraction results are in good agreement with the contour method. It was found that tensile stresses are present at the notch root and the free edge areas, and compressive stress is seen in the middle of the sample. The tensile stresses in the as-built condition resulted in a higher fatigue crack growth rate. After stress relieving by heat treatment, the tensile residual stress diminished by around 90%, resulting in decreased crack growth rate. The build direction was seen to affect the crack growth rate, although the trend was different between the as-built and stress-relieved conditions.

Recent advances in transformation optics

Abstract: Within the past a few years, transformation optics has emerged as a new research area, since it provides a general methodology and design tool for manipulating electromagnetic waves in a prescribed manner. Using transformation optics, researchers have demonstrated a host of striking phenomena and devices; many of which were only thought possible in science fiction. In this paper, we review the most recent advances in transformation optics. We focus on the theory, design, fabrication and characterization of transformation devices such as the carpet cloak, “Janus” lens and plasmonic cloak at optical frequencies, which allow routing light at the nanoscale. We also provide an outlook of the challenges and future directions in this fascinating area of transformation optics.

First Detection of sub-PeV Diffuse Gamma Rays from the Galactic Disk: Evidence for Ubiquitous Galactic Cosmic Rays beyond PeV Energies.

Abstract: We report, for the first time, the long-awaited detection of diffuse gamma rays with energies between 100 TeV and 1 PeV in the Galactic disk Particularly, all gamma rays above 398 TeV are observed apart from known TeV gamma-ray sources and compatible with expectations from the hadronic emission scenario in which gamma rays originate from the decay of π^{0}'s produced through the interaction of protons with the interstellar medium in the Galaxy This is strong evidence that cosmic rays are accelerated beyond PeV energies in our Galaxy and spread over the Galactic disk

Nonlinear optical selection rule based on valley-exciton locking in monolayer ws2

Abstract: Optical selection rules fundamentally determine the optical transitions between energy states in a variety of physical systems, from hydrogen atoms to bulk crystals such as gallium arsenide. These rules are important for optoelectronic applications such as lasers, energy-dispersive X-ray spectroscopy, and quantum computation. Recently, single-layer transition metal dichalcogenides have been found to exhibit valleys in momentum space with nontrivial Berry curvature and excitons with large binding energy. However, there has been little study of how the unique valley degree of freedom combined with the strong excitonic effect influences the nonlinear optical excitation. Here, we report the discovery of nonlinear optical selection rules in monolayer WS2, an important candidate for visible 2D optoelectronics because of its high quantum yield and large direct bandgap. We experimentally demonstrated this principle for second-harmonic generation and two-photon luminescence (TPL). Moreover, the circularly polarized TPL and the study of its dynamics evince a sub-ps interexciton relaxation (2p → 1s). The discovery of this new optical selection rule in a valleytronic 2D system not only considerably enhances knowledge in this area but also establishes a foundation for the control of optical transitions that will be crucial for valley optoelectronic device applications such as 2D valley-polarized THz sources with 2p–1s transitions, optical switches, and coherent control for quantum computing. An optical selection rule based on valley-exciton locking for nonlinear optical effects monolayer tungsten disulfide (WS2) is demonstrated. Optical selection rules derived from symmetry considerations control many light-based phenomena and applications. However, the effect of the combination of valley degree of freedom and strong excitonic effects on nonlinear optical excitation has not been extensively studied. Now, by considering energy valleys in momentum space, Xiang Zhang and co-workers at the University of California at Berkeley, have derived an optical selection rule for nonlinear optical effects of WS2, an important material for optoelectronic applications. They experimentally demonstrated the rule for second-harmonic generation and two-photon luminescence. This optical selection rule for a two-dimensional valleytronic system provides an important foundation for controlling optical transitions in applications of valley optoelectronics.

J/psi Elliptic Flow in Pb-Pb Collisions at root s(NN)=2.76 TeV

Abstract: We report on the first measurement of inclusive J/psi elliptic flow v(2) in heavy-ion collisions at the LHC. The measurement is performed with the ALICE detector in Pb-Pb collisions at root s(NN) = 2.76 TeV in the rapidity range 2.5 < y < 4.0. The dependence of the J/psi v(2) on the collision centrality and on the J/psi transverse momentum is studied in the range 0 <= p(T) < 10 GeV/c. For semicentral Pb-Pb collisions at root s(NN) = 2.76 TeV, an indication of nonzero v(2) is observed with a largest measured value of v(2) = 0.116 +/-0.046(stat) +/- 0.029(syst) for J/psi in the transverse momentum range 2 <= p(T) < 4 GeV/c. The elliptic flow measurement complements the previously reported ALICE results on the inclusive J/psi nuclear modification factor and favors the scenario of a significant fraction of J/psi production from charm quarks in a deconfined partonic phase.

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.