scispace - formally typeset
Search or ask a question
Author

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.


Papers
More filters
Journal ArticleDOI
TL;DR: The hybridization between the surface plasmon modes and waveguide modes allows efficient optical trapping of single dielectric nanoparticle with size of only several nanometers in the gap region, manifesting various optomechanical applications such as nanoscale optical tweezers.
Abstract: We demonstrate that in a hybrid plasmonic system the optical force exerted on a dielectric waveguide by a metallic substrateisenhancedbymorethan1orderofmagnitudecompared to the force between a photonic waveguide and a dielectric subs- trate. A nanoscale gap between the dielectric waveguide and the metallic substrate leads to deep subwavelength optical energy con- finement with ultralow mode propagation loss and hence results in the enhanced optical forces at low input optical power, as numeri- cally demonstrated by both Maxwell's stress tensor formalism and the coupled mode theory analysis. Moreover, the hybridization between the surface plasmon modes and waveguide modes allows efficient optical trapping of single dielectric nanoparticle with size of only several nanometers in the gap region, manifesting various optomechanical applications such as nanoscale optical tweezers.

210 citations

Journal ArticleDOI
TL;DR: The experimental generation and dynamic trajectory control of plasmonic Airy beams (PABs) are reported and it is shown that the ballistic motion of the PABs can be reconfigured in real time by either a computer addressed spatial light modulator or mechanical means.
Abstract: We report the experimental generation and dynamic trajectory control of plasmonic Airy beams (PABs). The PABs are created by directly coupling free-space Airy beams to surface plasmon polaritons through a grating coupler on a metal surface. We show that the ballistic motion of the PABs can be reconfigured in real time by either a computer addressed spatial light modulator or mechanical means.

206 citations

Journal ArticleDOI
TL;DR: The design, fabrication and characterization of optical hyperlens that can image sub-diffraction-limited objects in the far field are reported, based on an artificial anisotropic metamaterial with carefully designed hyperbolic dispersion composed of curved silver/alumina multilayers.
Abstract: We report here the design, fabrication and characterization of optical hyperlens that can image sub-diffraction-limited objects in the far field. The hyperlens is based on an artificial anisotropic metamaterial with carefully designed hyperbolic dispersion. We successfully designed and fabricated such a metamaterial hyperlens composed of curved silver/alumina multilayers. Experimental results demonstrate far-field imaging with resolution down to 125nm at 365nm working wavelength which is below the diffraction limit.

205 citations

Journal ArticleDOI
TL;DR: In this article, a two-dimensional cubic lattice consisting of thin metal wires, having wire diameter of 30 μm, lattice constant of 120 μm and wire length of 1 mm, was constructed using microstereolithography.
Abstract: Metamaterials, which contain engineered subwavelength microstructures, can be designed to have positive or negative e and μ at desired frequencies. In this letter, we demonstrate a metamaterial which has a “plasmonic” response to electromagnetic waves in the terahertz (THz) range. The sharp change of reflection and transmission at this plasma frequency makes the structure a high pass filter. The reflection response is characterized by Fourier transform infrared spectroscopy, and a plasma frequency at 0.7 THz is observed, which agrees with the theoretical calculation. The metamaterial is a two-dimensional cubic lattice consisting of thin metal wires, having wire diameter of 30 μm, lattice constant of 120 μm, and wire length of 1 mm. The microstereolithography technique is employed to fabricate the high-aspect-ratio lattice.

204 citations

Journal ArticleDOI
TL;DR: This work presents the first study of subwavelength discrete solitons in nonlinear metamaterials: nanoscaled periodic structures consisting of metal and nonlinear dielectric slabs, and expects these phenomena to inspire fundamental studies as well as potential applications of non linear metammaterials, particularly in subwa wavelength nonlinear optics.
Abstract: We present the first study of subwavelength discrete solitons in nonlinear metamaterials: nanoscaled periodic structures consisting of metal and nonlinear dielectric slabs. The solitons supported by such media result from a balance between tunneling of surface plasmon modes and nonlinear self-trapping. The dynamics in such systems, arising from the threefold interplay between periodicity, nonlinearity, and surface plasmon polaritons, is substantially different from that in conventional nonlinear dielectric waveguide arrays. We expect these phenomena to inspire fundamental studies as well as potential applications of nonlinear metamaterials, particularly in subwavelength nonlinear optics.

203 citations


Cited by
More filters
Proceedings ArticleDOI
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
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.

123,388 citations

Proceedings Article
04 Sep 2014
TL;DR: This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small 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.
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.

55,235 citations

Journal ArticleDOI
04 Mar 2011-Cell
TL;DR: Recognition of the widespread applicability of these concepts will increasingly affect the development of new means to treat human cancer.

51,099 citations

Posted Content
TL;DR: This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth.
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.

44,703 citations

Proceedings ArticleDOI
07 Jun 2015
TL;DR: Inception as mentioned in this paper is a deep convolutional neural network architecture that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
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.

40,257 citations