arXiv subjects

Han Zhang

Publications and source records attributed to Han Zhang.

At least 469 records · Page 26Linked to original sources

arXiv2018

Over 800% Efficiency Enhancement of Solution-Processed All-Inorganic Quantum-Dot Light Emitting Diodes with an Ultrathin Alumina Passivating Layer

The use of robust, inorganic charge-transport materials is always desired in quantum-dot light emitting diodes (QLEDs) because they are expected to allow higher stability and less cost than that of organic counterparts. Here we report an all-inorganic QLED with excellent efficiency by modifying the solution-processed NiO (s-NiO) surface with an ultrathin Al2O3 passivating layer. The localized electric field induced by nickel oxyhydroxide (NiOOH) is estimated to be ~ 70 MV/cm at a distance of 6 nm from the surface of s-NiO layer. Both transient resolution photoluminescence (TRPL) and X-Ray photoelectron spectroscopy (XPS) measurements demonstrate that the Al2O3 passivating layer can effectively passivate the NiOOH on the s-NiO surface, hence suppressing the exciton quenching. As a result, over 800% efficiency enhancement up to 34.1 cd/A (8.1%) for the current efficiency (external quantum efficiency, EQE) of the QLEDs is achieved. To the best of our knowledge, this is the best-performing all-inorganic QLED so far.

physics.app-ph↗

arXiv2018

Atomic inner-shell radiation seeded free-electron lasers

In order to effectively improve the output quality of X-ray free electron laser (XFEL), we theoretically propose an XFEL scheme seeded by atomic inner-shell laser. As well known, an atomic inner-shell laser based on neutral atoms and pumped by an XFEL has been experimentally demonstrated, which produced sub-femtosecond X-ray pulses with increased temporal coherence. It shows that, by using the inner-shell laser as a seed to modulate the electron bunch, very stable and almost fully-coherent short-wavelength XFEL pulses can be generated. The proposed scheme holds promising prospects in X-ray wavelengths, and even shorter.

physics.acc-ph↗

arXiv2018

Improving GANs Using Optimal Transport

We present Optimal Transport GAN (OT-GAN), a variant of generative adversarial nets minimizing a new metric measuring the distance between the generator distribution and the data distribution. This metric, which we call mini-batch energy distance, combines optimal transport in primal form with an energy distance defined in an adversarially learned feature space, resulting in a highly discriminative distance function with unbiased mini-batch gradients. Experimentally we show OT-GAN to be highly stable when trained with large mini-batches, and we present state-of-the-art results on several popular benchmark problems for image generation.

cs.LG↗

arXiv2017

AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks

In this paper, we propose an Attentional Generative Adversarial Network (AttnGAN) that allows attention-driven, multi-stage refinement for fine-grained text-to-image generation. With a novel attentional generative network, the AttnGAN can synthesize fine-grained details at different subregions of the image by paying attentions to the relevant words in the natural language description. In addition, a deep attentional multimodal similarity model is proposed to compute a fine-grained image-text matching loss for training the generator. The proposed AttnGAN significantly outperforms the previous state of the art, boosting the best reported inception score by 14.14% on the CUB dataset and 170.25% on the more challenging COCO dataset. A detailed analysis is also performed by visualizing the attention layers of the AttnGAN. It for the first time shows that the layered attentional GAN is able to automatically select the condition at the word level for generating different parts of the image.

cs.CV↗

arXiv2017

Skyrmion dynamics in a frustrated ferromagnetic film and current-induced helicity locking-unlocking transition

The helicity-orbital coupling is an intriguing feature of magnetic skyrmions in frustrated magnets. Here, we explore the skyrmion dynamics in a frustrated magnet based on the $J_{1}$-$J_{2}$-$J_{3}$ classical Heisenberg model explicitly by including the dipole-dipole interaction. The skyrmion energy acquires a helicity dependence due to the dipole-dipole interaction, resulting in the current-induced translational motion with a fixed helicity. The lowest energy states are the degenerate Bloch-type states, which can be used for building the binary memory. By increasing the driving current, the helicity locking-unlocking transition occurs, where the translational motion changes to the rotational motion. Furthermore, we demonstrate that two skyrmions can spontaneously form a bound state. The separation of the bound state forced by a driving current is also studied. In addition, we show the annihilation of a pair of skyrmion and antiskyrmion. Our results reveal the distinctive frustrated skyrmions may enable viable applications in topological magnetism.

cond-mat.mes-hall↗

arXiv2017

Dynamic cavity effect in Topological Insulator Bi2Te3 based Passive Q-switched Solid State Laser

Dynamic cavity effect of a Topological Insulator Bi2Te3 based Passive Q-switched (QS) solid state laser was investigated. With a saturation intensity as low as a few kW, Bi2Te3 processed by hydrothermal exfoliation serves a sensitive saturable absorber to initiate Q-switching. Without material damage, a transition from QS state to CW state was observed when the pumping strength increases which was explained by the modulation instability spectrum of laser rate equation.

physics.optics↗

arXiv2017

Consensus Control for Linear Systems with Optimal Energy Cost

In this paper, we design an optimal energy cost controller for linear systems asymptotic consensus given the topology of the graph. The controller depends only on relative information of the agents. Since finding the control gain for such controller is hard, we focus on finding an optimal controller among a classical family of controllers which is based on Algebraic Riccati Equation (ARE) and guarantees asymptotic consensus. Through analysis, we find that the energy cost is bounded by an interval and hence we minimize the upper bound. In order to do that, there are two classes of variables that need to be optimized: the control gain and the edge weights of the graph and are hence designed from two perspectives. A suboptimal control gain is obtained by choosing $Q=0$ in the ARE. Negative edge weights are allowed, and the problem is formulated as a Semi-definite Programming (SDP) problem. Having negative edge weights means that "competitions" between the agents are allowed. The motivation behind this setting is to have a better system performance. We provide a different proof from the angle of optimization and show that the lowest control energy cost is reached when the graph is complete and with equal edge weights. Furthermore, two sufficient conditions for the existence of negative optimal edge weights realization are given.

math.OC↗

arXiv2017

Link the head to the "beak": Zero Shot Learning from Noisy Text Description at Part Precision

In this paper, we study learning visual classifiers from unstructured text descriptions at part precision with no training images. We propose a learning framework that is able to connect text terms to its relevant parts and suppress connections to non-visual text terms without any part-text annotations. For instance, this learning process enables terms like "beak" to be sparsely linked to the visual representation of parts like head, while reduces the effect of non-visual terms like "migrate" on classifier prediction. Images are encoded by a part-based CNN that detect bird parts and learn part-specific representation. Part-based visual classifiers are predicted from text descriptions of unseen visual classifiers to facilitate classification without training images (also known as zero-shot recognition). We performed our experiments on CUBirds 2011 dataset and improves the state-of-the-art text-based zero-shot recognition results from 34.7\% to 43.6\%. We also created large scale benchmarks on North American Bird Images augmented with text descriptions, where we also show that our approach outperforms existing methods. Our code, data, and models are publically available.

cs.CV↗

arXiv2017

Mid-Infrared Polarization of Herbig Ae/Be Discs

We measured mid-infrared polarization of protoplanetary discs to gain new insight into their magnetic fields. Using CanariCam at the 10.4 m Gran Telescopio Canarias, we detected linear polarization at 8.7, 10.3, and 12.5 $μ$m from discs around eight Herbig Ae/Be stars and one T-Tauri star. We analyzed polarimetric properties of each object to find out the most likely interpretation of the data. While the observed mid-infrared polarization from most objects is consistent with polarized emission and/or absorption arising from aligned dust particles, we cannot rule out polarization due to dust scattering for a few objects in our sample. For those objects for which polarization can be explained by polarized emission and/or absorption, we examined how the derived magnetic field structure correlates with the disc position angle and inclination. We found no preference for a certain type of magnetic field. Instead, various configurations (toroidal, poloidal, or complex) are inferred from the observations. The detection rate (64 per cent) of polarized mid-infrared emission and/or absorption supports the expectation that magnetic fields and suitable conditions for grain alignment are common in protoplanetary discs around Herbig Ae/Be stars.

astro-ph.EP↗

arXiv2017

StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks

Synthesizing high-quality images from text descriptions is a challenging problem in computer vision and has many practical applications. Samples generated by existing text-to-image approaches can roughly reflect the meaning of the given descriptions, but they fail to contain necessary details and vivid object parts. In this paper, we propose Stacked Generative Adversarial Networks (StackGAN) to generate 256x256 photo-realistic images conditioned on text descriptions. We decompose the hard problem into more manageable sub-problems through a sketch-refinement process. The Stage-I GAN sketches the primitive shape and colors of the object based on the given text description, yielding Stage-I low-resolution images. The Stage-II GAN takes Stage-I results and text descriptions as inputs, and generates high-resolution images with photo-realistic details. It is able to rectify defects in Stage-I results and add compelling details with the refinement process. To improve the diversity of the synthesized images and stabilize the training of the conditional-GAN, we introduce a novel Conditioning Augmentation technique that encourages smoothness in the latent conditioning manifold. Extensive experiments and comparisons with state-of-the-arts on benchmark datasets demonstrate that the proposed method achieves significant improvements on generating photo-realistic images conditioned on text descriptions.

cs.CV↗

arXiv2017

Detection of Polarized Infrared Emission by Polycyclic Aromatic Hydrocarbons in the MWC 1080 Nebula

Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous in astrophysical environments, as revealed by their pronounced emission features at 3.3, 6.2, 7.7, 8.6, 11.3, and 12.7 $μ$m commonly ascribed to the C--H and C--C vibrational modes. Although these features have long been predicted to be polarized, previous searches for PAH polarization led to null or, at best, tentative detections. Here we report the definite detection of polarized PAH emission at 11.3 $μ$m in the nebula associated with the Herbig Be star MWC 1080. We measure a polarization degree of 1.9$\pm$0.2\%, which is unexpectedly high compared to models. This poses a challenge in the current understanding of the alignment of PAHs, which is required to polarize the PAH emission but thought to be substantially suppressed. PAH alignment with a magnetic field via a resonance paramagnetic relaxation process may account for such a high level of polarization.

astro-ph.GA↗

arXiv2017

SegAN: Adversarial Network with Multi-scale $L_1$ Loss for Medical Image Segmentation

Inspired by classic generative adversarial networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale $L_1$ loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. In our SegAN framework, the segmentor and critic networks are trained in an alternating fashion in a min-max game: The critic takes as input a pair of images, (original_image $*$ predicted_label_map, original_image $*$ ground_truth_label_map), and then is trained by maximizing a multi-scale loss function; The segmentor is trained with only gradients passed along by the critic, with the aim to minimize the multi-scale loss function. We show that such a SegAN framework is more effective and stable for the segmentation task, and it leads to better performance than the state-of-the-art U-net segmentation method. We tested our SegAN method using datasets from the MICCAI BRATS brain tumor segmentation challenge. Extensive experimental results demonstrate the effectiveness of the proposed SegAN with multi-scale loss: on BRATS 2013 SegAN gives performance comparable to the state-of-the-art for whole tumor and tumor core segmentation while achieves better precision and sensitivity for Gd-enhance tumor core segmentation; on BRATS 2015 SegAN achieves better performance than the state-of-the-art in both dice score and precision.

cs.CV↗

arXiv2017

Tuning of Interlayer Coupling in Large-Area Graphene/WSe2 van der Waals Heterostructure via Ion Irradiation: Optical Evidences and Photonic Applications

Van der Waals (vdW) heterostructures are receiving great attentions due to their intriguing properties and potentials in many research fields. The flow of charge carriers in vdW heterostructures can be efficiently rectified by the inter-layer coupling between neighboring layers, offering a rich collection of functionalities and a mechanism for designing atomically thin devices. Nevertheless, non-uniform contact in larger-area heterostructures reduces the device efficiency. In this work, ion irradiation had been verified as an efficient technique to enhance the contact and interlayer coupling in the newly developed graphene/WSe2 hetero-structure with a large area of 10 mm x 10 mm. During the ion irradiation process, the morphology of monolayer graphene had been modified, promoting the contact with WSe2. Experimental evidences of the tunable interlayer electron transfer are displayed by investigation of photoluminescence and ultrafast absorption of the irradiated heterostructure. Besides, we have found that in graphene/WSe2 heterostructure, graphene serves as a fast channel for the photo-excited carriers to relax in WSe2, and the nonlinear absorption of WSe2 could be effectively tuned by the carrier transfer process in graphene, enabling specific optical absorption of the heterostructure in comparison with separated graphene or WSe2. On the basis of these new findings, by applying the ion beam modified graphene/WSe2 heterostructure as a saturable absorber, Q-switched pulsed lasing with optimized performance has been realized in a Nd:YAG waveguide cavity. This work paves the way towards developing novel devices based on large-area heterostructures by using ion beam irradiation.

cond-mat.mes-hall↗

arXiv2017

Bandwidth broadening of X-ray free electron laser pulses with the natural gradient of planar undulator

Besides the target to pursue the narrow bandwidth X-ray pulses, the large bandwidth free-electron laser pulses are also strongly demanded to satisfy a wide range of scientific user experiments. In this paper, using the transversely tilt beam enabled by deflecting cavity and/or corrugated structure, the potential of large bandwidth X-ray free-electron lasers generation with the natural gradient of the planar undulator are discussed. Simulations confirm the theoretical prediction, and X-ray free-electron laser bandwidth indicates an increase of one order of magnitude with the optimized parameters.

physics.acc-ph↗

arXiv2017

One-Pass Error Bounded Trajectory Simplification

Nowadays, various sensors are collecting, storing and transmitting tremendous trajectory data, and it is known that raw trajectory data seriously wastes the storage, network band and computing resource. Line simplification (LS) algorithms are an effective approach to attacking this issue by compressing data points in a trajectory to a set of continuous line segments, and are commonly used in practice. However, existing LS algorithms are not sufficient for the needs of sensors in mobile devices. In this study, we first develop a one-pass error bounded trajectory simplification algorithm (OPERB), which scans each data point in a trajectory once and only once. We then propose an aggressive one-pass error bounded trajectory simplification algorithm (OPERB-A), which allows interpolating new data points into a trajectory under certain conditions. Finally, we experimentally verify that our approaches (OPERB and OPERB-A) are both efficient and effective, using four real-life trajectory datasets.

cs.DB↗

arXiv2016

On Hybrid Pilot for Channel Estimation in Massive MIMO Uplink

This paper introduces a hybrid pilot-aided channel estimation technique for mitigating the effect of pilot contamination for the uplink of multi-cell multiuser massive MIMO systems. The proposed hybrid pilot is designed such that it enjoys the complementary advantages between time-multiplexed (TM) pilot and time-superimposed (TS) pilot, and thereby, allows superior solution to the conventional pilot schemes. We mathematically characterize the impact of hybrid pilot on the massive MIMO uplink by deriving a closed-form approximation for the uplink achievable rate. In large-number-of-antennas regime, we obtain the asymptotically optimal solution for hybrid pilot by jointly designing the TM pilot and the TS pilot. It is shown that either TM pilot or TS pilot has the advantages for large frame-size and limited frame-size transmission, respectively, while the hybrid pilot scheme can offer a superior performance to that employing either TM pilot or TS pilot. Numerical results demonstrate the effectiveness of the proposed design.

cs.IT↗

arXiv2016

Size-Dependent Structural Phase Transitions in SrTiO3 Nanoparticles

Understanding the structural phase diagram of nano scale SrTiO3 has important implications on the basic physics and applications of the general class of transition metal oxide perovskites. Pressure dependent structural measurements on monodispersed nanoscale SrTiO3 samples with average diameters of 10 to ~80 nm were conducted. A robust pressure independent polar structure was detected in the 10 nm sample for pressures of up to 13 GPa while a size dependent cubic to tetragonal transition occurs (at P = Pc) for larger particle sizes. The results suggest that the growth of ~10 nm STO particles on substrates with large lattice mismatch will not alter the polar state of the system for a large range of strain values, possibly enabling device use.

cond-mat.mtrl-sci↗

arXiv2016

The Mid-Infrared Polarization of the Herbig Ae Star WL 16: An Interstellar Origin?

We present high-resolution (0".4) mid-infrared (mid-IR) polarimetric images and spectra of WL 16, a Herbig Ae star at a distance of 125 pc. WL 16 is surrounded by a protoplanetary disk of $\sim$ 900 AU in diameter, making it one of the most extended Herbig Ae/Be disks as seen in the mid-IR. The star is behind, or embedded in, the $ρ$ Ophiuchus molecular cloud, and obscured by 28 magnitudes of extinction at optical wavelengths by the foreground cloud. Mid-IR polarization of WL 16, mainly arises from aligned elongated dust grains present along the line of sight, suggesting a uniform morphology of polarization vectors with an orientation of 33\degr (East from North) and a polarization fraction of $\sim$ 2.0\%. This orientation is consistent with previous polarimetric surveys in the optical and near-IR bands to probe large-scale magnetic fields in the Ophiuchus star formation region, indicating that the observed mid-IR polarization toward WL 16 is produced by the dichroic absorption of magnetically aligned foreground dust grains by a uniform magnetic field. Using polarizations of WL 16 and Elias 29, a nearby polarization standard star, we constrain the polarization efficiency, \textit{$p_{10.3}/A_{10.3}$}, for the dust grains in the $ρ$ Ophiuchus molecular cloud to be $\sim$ 1.0\% mag$^{-1}$. WL 16 has polycyclic aromatic hydrocarbon (PAH) emission features detected at 8.6, 11.2, 12.0, and 12.7 $μ$m by our spectroscopic data, and we find an anti-correlation between the PAH surface brightness and the PAH ionization fraction between the NW and SW sides of the disk.

astro-ph.EP↗