arXiv subjects

Kang Wang

Publications and source records attributed to Kang Wang.

105 records · Page 6Linked to original sources

arXiv2021

Self-stabilizing exchange-mediated spin transport

Long-range spin transport in magnetic systems can be achieved by means of exchange-mediated spin textures with robust topological winding -- a phenomenon referred to as spin superfluidity. Its experimental signatures have been discussed in antiferromagnets which are nearly free of dipolar interaction. However, in ferromagnets, which possess non-negligible dipole fields, realization of such spin transport has remained a challenge. Using micromagnetic simulations, we investigate coherent exchange-mediated spin transport in extended thin ferromagnetic films. We uncover a two-fluid state, in which the long-range spin transport by spin textures co-exists with spin waves, as well as a soliton-screened spin transport regime at high spin injection biases. Both states are associated with distinct spin texture reconstructions near the spin injection region and sustain spin transport over large distances.

cond-mat.other↗

arXiv2021

Shear based gap control in 2D photonic quasicrystals of dielectric cylinders

2D dielectric photonic quasicrystals can be designed to show isotropic band gaps. The system here studied is a quasiperiodic lattice made of silicon dielectric cylinders arranged as periodic unit cell based on a decagonal approximant of a quasiperiodic Penrose lattice. We analyze the bulk properties of the resulting lattice as well as the bright states excited in the gap which correspond to localized resonances of the electromagnetic field in specific cylinder clusters of the lattice. Then we introduce a controlled shear deformation which breaks the decagonal symmetry and evaluate the width reduction of the gap together with the evolution of the resonances, for all shear values compatible with physical constraints (cylinder collision). The gap is reduced up to a 18.5 % while different states change their frequency in different ways. Realistic analysis about the actual transmission of the electromagnetic radiation, often missing in the literature, have been performed for a finite slice of the proposed quasicrystals structure. Two calculation procedures based on MIT Photonic Bands (MPB) and Finite Integration Technique (FIT) are used for the bulk and the finite structures finding an extremely good agreement among them.

physics.optics↗

arXiv2021

Towards Accurate RGB-D Saliency Detection with Complementary Attention and Adaptive Integration

Saliency detection based on the complementary information from RGB images and depth maps has recently gained great popularity. In this paper, we propose Complementary Attention and Adaptive Integration Network (CAAI-Net), a novel RGB-D saliency detection model that integrates complementary attention based feature concentration and adaptive cross-modal feature fusion into a unified framework for accurate saliency detection. Specifically, we propose a context-aware complementary attention (CCA) module, which consists of a feature interaction component, a complementary attention component, and a global-context component. The CCA module first utilizes the feature interaction component to extract rich local context features. The resulting features are then fed into the complementary attention component, which employs the complementary attention generated from adjacent levels to guide the attention at the current layer so that the mutual background disturbances are suppressed and the network focuses more on the areas with salient objects. Finally, we utilize a specially-designed adaptive feature integration (AFI) module, which sufficiently considers the low-quality issue of depth maps, to aggregate the RGB and depth features in an adaptive manner. Extensive experiments on six challenging benchmark datasets demonstrate that CAAI-Net is an effective saliency detection model and outperforms nine state-of-the-art models in terms of four widely-used metrics. In addition, extensive ablation studies confirm the effectiveness of the proposed CCA and AFI modules.

cs.CV↗

arXiv2020

DeepOpht: Medical Report Generation for Retinal Images via Deep Models and Visual Explanation

In this work, we propose an AI-based method that intends to improve the conventional retinal disease treatment procedure and help ophthalmologists increase diagnosis efficiency and accuracy. The proposed method is composed of a deep neural networks-based (DNN-based) module, including a retinal disease identifier and clinical description generator, and a DNN visual explanation module. To train and validate the effectiveness of our DNN-based module, we propose a large-scale retinal disease image dataset. Also, as ground truth, we provide a retinal image dataset manually labeled by ophthalmologists to qualitatively show, the proposed AI-based method is effective. With our experimental results, we show that the proposed method is quantitatively and qualitatively effective. Our method is capable of creating meaningful retinal image descriptions and visual explanations that are clinically relevant.

cs.CV↗

arXiv2020

Efficient T2 mapping with Blip-up/down EPI and gSlider-SMS (T2-BUDA-gSlider)

Purpose: To rapidly obtain high isotropic-resolution T2 maps with whole-brain coverage and high geometric fidelity. Methods: A T2 blip-up/down echo planar imaging (EPI) acquisition with generalized Slice-dithered enhanced resolution (T2-BUDA-gSlider) is proposed. A radiofrequency (RF)-encoded multi-slab spin-echo EPI acquisition with multiple echo times (TEs) was developed to obtain high SNR efficiency with reduced repetition time (TR). This was combined with an interleaved 2-shot EPI acquisition using blip-up/down phase encoding. An estimated field map was incorporated into the joint multi-shot EPI reconstruction with a structured low rank constraint to achieve distortion-free and robust reconstruction for each slab without navigation. A Bloch simulated subspace model was integrated into gSlider reconstruction and utilized for T2 quantification. Results: In vivo results demonstrated that the T2 values estimated by the proposed method were consistent with gold standard spin-echo acquisition. Compared to the reference 3D fast spin echo (FSE) images, distortion caused by off-resonance and eddy current effects were effectively mitigated. Conclusion: BUDA-gSlider SE-EPI acquisition and gSlider-subspace joint reconstruction enabled distortion-free whole-brain T2 mapping in 2 min at ~1 mm3 isotropic resolution, which could bring significant benefits to related clinical and neuroscience applications.

physics.med-ph↗

arXiv2020

Improving Accent Conversion with Reference Encoder and End-To-End Text-To-Speech

Accent conversion (AC) transforms a non-native speaker's accent into a native accent while maintaining the speaker's voice timbre. In this paper, we propose approaches to improving accent conversion applicability, as well as quality. First of all, we assume no reference speech is available at the conversion stage, and hence we employ an end-to-end text-to-speech system that is trained on native speech to generate native reference speech. To improve the quality and accent of the converted speech, we introduce reference encoders which make us capable of utilizing multi-source information. This is motivated by acoustic features extracted from native reference and linguistic information, which are complementary to conventional phonetic posteriorgrams (PPGs), so they can be concatenated as features to improve a baseline system based only on PPGs. Moreover, we optimize model architecture using GMM-based attention instead of windowed attention to elevate synthesized performance. Experimental results indicate when the proposed techniques are applied the integrated system significantly raises the scores of acoustic quality (30$\%$ relative increase in mean opinion score) and native accent (68$\%$ relative preference) while retaining the voice identity of the non-native speaker.

cs.CL↗

arXiv2020

Quantized Auger Recombination of Polaronic Self-trapped Excitons in Bulk Iron Oxide

The Auger recombination in bulk semiconductors can depopulate the charge carriers in a non-radiative way, which, fortunately, only has detrimental impact on optoelectronic device performance under the condition of high carrier density because the restriction arising from concurrent momentum and energy conservation limits the Auger rate. Here, we surprisingly found that the Auger recombination in bulk Fe2O3 films was more efficient than narrow-bandgap high-mobility semiconductors that were supposed to have much higher Auger rate constants than metal oxides. The Auger process in Fe2O3 was ascribed to the Coulombically coupled self-trapped excitons (STEs), which was enhanced by the relaxation of momentum conservation because of the strong spatial localization of these STEs. Furthermore, due to this localization effect the kinetic traces of the STE annihilation for different STE densities exhibited characteristics of quantized Auger recombination, and we demonstrated that these traces could be simultaneously modeled by taking into account the quantized Auger rates.

physics.chem-ph↗

arXiv2018

Structures and Diffusion of Two Dimensional Dusty Plasmas on One Dimensional Periodic Substrates

Using numerical simulations, we examine the structure and diffusion of a two-dimensional dusty plasma (2DDP) in the presence of a one-dimensional periodic substrate (1DPS) as a function of increasing substrate strength. Both the pair correlation function perpendicular to the substrate modulation and the mean squared displacement (MSD) of dust particles are calculated. It is found that both the structure and dynamics of 2DDP exhibit strong anisotropic effects, due to the applied 1DPS. As the substrate strength increases from 0, the structure order of dusty plasma along each potential well of 1DPS increases first probably due to the competition between the inter-particle interactions and the particle-substrate interactions, and then decreases gradually, which may be due to the reduced dimensionality and the enhanced fluctuations. The obtained MSD along potential wells of 1DPS clearly shows three processes of diffusion in our studied 2DDP. Between the initial ballistic and finally diffusive motion, there is the intermediate sub-diffusion discovered here, which may result from the substrate-induced distortion of the caging dynamics.

physics.plasm-ph↗

arXiv2017

Summing coincidence in rare event gamma-ray measurements under an ultra-low background environment

A Monte Carlo method based on the GEANT4 toolkit has been developed to correct the full-energy peak (FEP) efficiencies of a high purity germanium (HPGe) detector equipped with a low background shielding system, and moreover evaluated using summing peaks in a numerical way. It is found that the FEP efficiencies of $^{60}$Co, $^{133}$Ba and $^{152}$Eu can be improved up to 18\% by taking the calculated true summing \mbox{coincidence} factors (TSCFs) correction into account. Counts of summing coincidence $γ$ peaks in the spectrum of $^{152}$Eu can be well reproduced using the corrected efficiency curve within an accuracy of 3\%.

physics.ins-det↗

arXiv2017

Simulation Experiment of BCI Based on Imagined Speech EEG Decoding

Brain Computer Interface (BCI) can help patients of neuromuscular diseases restore parts of the movement and communication abilities that they have lost. Most of BCIs rely on mapping brain activities to device instructions, but limited number of brain activities decides the limited abilities of BCIs. To deal with the problem of limited ablility of BCI, this paper verified the feasibility of constructing BCI based on decoding imagined speech electroencephalography (EEG). As sentences decoded from EEG can have rich meanings, BCIs based on EEG decoding can achieve numerous control instructions. By combining a modified EEG feature extraction mehtod with connectionist temporal classification (CTC), this paper simulated decoding imagined speech EEG using synthetic EEG data without help of speech signal. The performance of decoding model over synthetic data to a certain extent demonstrated the feasibility of constructing BCI based on imagined speech brain signal.

cs.HC↗

arXiv2017

Quantum speed limit for a relativistic electron in the noncommutative phase space

The influence of the noncommutativity on the average speed of a relativistic electron interacting with a uniform magnetic field within the minimum evolution time is investigated. We find that it is possible for the wave packet of the electron to travel faster than the speed of light in vacuum because of the noncommutativity. It suggests that due to the noncommutativity, Lorentz invariance is violated in the relativistic quantum mechanics region.

hep-th↗

arXiv2016

Optimal width of quasi-crystalline slabs of dielectric cylinders to light transmission contrast

Light confinement induced by resonant states in aperiodic photonic structures are interesting for many applications. A particular case of these resonances can be found in 2D quasi-crystalline arrangements of dielectric cylinders. These systems present a rather isotropic band gap as well as isolated in-gap photonic states (as a result of spatially localized resonances). These states are built by high symmetry polygonal clusters that can be regarded as photonic molecules. In this paper we study the transmission properties of a slab of glass cylinders arranged in approximants of decagonal quasi-crystalline structure. In particular, we investigate the influence of the slab width in the transmission contrast between the states and the gap. The study is both experimental and numerical in the microwave regime. We find that the best transmission contrast is found for a width of around 3 times the radiation wavelength. The transmission at the band gap region is mediated by the resonances of the photonic molecules. If the samples are thin enough they become transparent except around a resonance of the photonic molecule which reflects the incoming light. In this case the behavior is reminiscent of an absorbing molecule.

physics.optics↗

arXiv2015

Sense Beyond Expressions: Cuteness

With the development of Internet culture, cuteness has become a popular concept. Many people are curious about what factors making a person look cute. However, there is rare research to answer this interesting question. In this work, we construct a dataset of personal images with comprehensively annotated cuteness scores and facial attributes to investigate this high-level concept in depth. Based on this dataset, through an automatic attributes mining process, we find several critical attributes determining the cuteness of a person. We also develop a novel Continuous Latent Support Vector Machine (C-LSVM) method to predict the cuteness score of one person given only his image. Extensive evaluations validate the effectiveness of the proposed method for cuteness prediction.

cs.CV↗

arXiv2008

Quantum Spin Hall Effect in Inverted Type II Semiconductors

The quantum spin Hall (QSH) state is a topologically non-trivial state of quantum matter which preserves time-reversal symmetry; it has an energy gap in the bulk, but topologically robust gapless states at the edge. Recently, this novel effect has been predicted and observed in HgTe quantum wells. In this work we predict a similar effect arising in Type-II semiconductor quantum wells made from InAs/GaSb/AlSb. Because of a rare band alignment the quantum well band structure exhibits an "inverted" phase similar to CdTe/HgTe quantum wells, which is a QSH state when the Fermi level lies inside the gap. Due to the asymmetric structure of this quantum well, the effects of inversion symmetry breaking and inter-layer charge transfer are essential. By standard self-consistent calculations, we show that the QSH state persists when these corrections are included, and a quantum phase transition between the normal insulator and the QSH phase can be electrically tuned by the gate voltage.

cond-mat.mes-hall↗

arXiv1999

Electron Spin Resonance Transistors for Quantum Computing in Silicon-Germanium Heterostructures

We apply the full power of modern electronic band structure engineering and epitaxial heterostructures to design a transistor that can sense and control a single donor electron spin. Spin resonance transistors may form the technological basis for quantum information processing. One and two qubit operations are performed by applying a gate bias. The bias electric field pulls the electron wave function away from the dopant ion into layers of different alloy composition. Owing to the variation of the g-factor (Si:g=1.995, Ge:g=1.563), this displacement changes the spin Zeeman energy, allowing single-qubit operations. By displacing the electron even further, the overlap with neighboring qubits is affected, which allows two-qubit operations. Certain Silicon-Germanium alloys allow a qubit spacing as large as 200 nm, which is well within the capabilities of current lithographic techniques. We discuss manufacturing limitations and issues regarding scaling up to a large size computer.

quant-ph↗