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.

Switching terahertz waves with gate-controlled active graphene metamaterials

Abstract: The extraordinary electronic properties of graphene provided the main thrusts for the rapid advance of graphene electronics In photonics, the gate-controllable electronic properties of graphene provide a route to efficiently manipulate the interaction of photons with graphene, which has recently sparked keen interest in graphene plasmonics However, the electro-optic tuning capability of unpatterned graphene alone is still not strong enough for practical optoelectronic applications owing to its non-resonant Drude-like behaviour Here, we demonstrate that substantial gate-induced persistent switching and linear modulation of terahertz waves can be achieved in a two-dimensional metamaterial, into which an atomically thin, gated two-dimensional graphene layer is integrated The gate-controllable light-matter interaction in the graphene layer can be greatly enhanced by the strong resonances of the metamaterial Although the thickness of the embedded single-layer graphene is more than six orders of magnitude smaller than the wavelength (<λ/1,000,000), the one-atom-thick layer, in conjunction with the metamaterial, can modulate both the amplitude of the transmitted wave by up to 47% and its phase by 322° at room temperature More interestingly, the gate-controlled active graphene metamaterials show hysteretic behaviour in the transmission of terahertz waves, which is indicative of persistent photonic memory effects

Metamaterials: a new frontier of science and technology

Abstract: Metamaterials, artificial composite structures with exotic material properties, have emerged as a new frontier of science involving physics, material science, engineering and chemistry. This critical review focuses on the fundamentals, recent progresses and future directions in the research of electromagnetic metamaterials. An introduction to metamaterials followed by a detailed elaboration on how to design unprecedented electromagnetic properties of metamaterials is presented. A number of intriguing phenomena and applications associated with metamaterials are discussed, including negative refraction, sub-diffraction-limited imaging, strong optical activities in chiral metamaterials, interaction of meta-atoms and transformation optics. Finally, we offer an outlook on future directions of metamaterials research including but not limited to three-dimensional optical metamaterials, nonlinear metamaterials and “quantum” perspectives of metamaterials (142 references).

Genome-wide RNAi screen of Ca(2+) influx identifies genes that regulate Ca(2+) release-activated Ca(2+) channel activity.

Abstract: Recent studies by our group and others demonstrated a required and conserved role of Stim in store-operated Ca(2+) influx and Ca(2+) release-activated Ca(2+) (CRAC) channel activity. By using an unbiased genome-wide RNA interference screen in Drosophila S2 cells, we now identify 75 hits that strongly inhibited Ca(2+) influx upon store emptying by thapsigargin. Among these hits are 11 predicted transmembrane proteins, including Stim, and one, olf186-F, that upon RNA interference-mediated knockdown exhibited a profound reduction of thapsigargin-evoked Ca(2+) entry and CRAC current, and upon overexpression a 3-fold augmentation of CRAC current. CRAC currents were further increased to 8-fold higher than control and developed more rapidly when olf186-F was cotransfected with Stim. olf186-F is a member of a highly conserved family of four-transmembrane spanning proteins with homologs from Caenorhabditis elegans to human. The endoplasmic reticulum (ER) Ca(2+) pump sarco-/ER calcium ATPase (SERCA) and the single transmembrane-soluble N-ethylmaleimide-sensitive (NSF) attachment receptor (SNARE) protein Syntaxin5 also were required for CRAC channel activity, consistent with a signaling pathway in which Stim senses Ca(2+) depletion within the ER, translocates to the plasma membrane, and interacts with olf186-F to trigger CRAC channel activity.

First observation of the Greisen-Zatsepin-Kuzmin suppression.

Abstract: The High Resolution Fly's Eye (HiRes) experiment has observed the Greisen-Zatsepin-Kuzmin suppression (called the GZK cutoff) with a statistical significance of five standard deviations. HiRes' measurement of the flux of ultrahigh energy cosmic rays shows a sharp suppression at an energy of $6\ifmmode\times\else\texttimes\fi{}{10}^{19}\text{ }\text{ }\mathrm{eV}$, consistent with the expected cutoff energy. We observe the ankle of the cosmic-ray energy spectrum as well, at an energy of $4\ifmmode\times\else\texttimes\fi{}{10}^{18}\text{ }\text{ }\mathrm{eV}$. We describe the experiment, data collection, and analysis and estimate the systematic uncertainties. The results are presented and the calculation of the statistical significance of our observation is described.

Projection micro-stereolithography using digital micro-mirror dynamic mask

Abstract: We present in this paper the development of a high-resolution projection micro-stereolithography (PμSL) process by using the Digital Micromirror Device (DMD™, Texas Instruments) as a dynamic mask. This unique technology provides a parallel fabrication of complex three-dimensional (3D) microstructures used for micro electro-mechanical systems (MEMS). Based on the understanding of underlying mechanisms, a process model has been developed with all critical parameters obtained from the experimental measurement. By coupling the experimental measurement and the process model, the photon-induced curing behavior of the resin has been quantitatively studied. The role of UV doping has been thereafter justified, as it can effectively reduce the curing depth without compromising the chemical property of the resin. The fabrication of complex 3D microstructures, such as matrix, and micro-spring array, with the smallest feature of 0.6 μm, has been demonstrated.

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.