arXiv subjects

Xiao Zhang

Publications and source records attributed to Xiao Zhang.

At least 469 records · Page 26Linked to original sources

arXiv2019

Entangled multi-component 4D quantum Hall states from photonic crystal defects

Recently, there has been a drive towards the realization of topological phases beyond conventional electronic materials, including phases defined in more than three dimensions. We propose a versatile and experimentally realistic approach of realizing a large variety of multi-component topological phases in 2D photonic crystals with quasi-periodically modulated defects. With a length scale introduced by a background resonator lattice, the defects are found to host various effective orbitals of $s$, $p$ and $d$-type symmetries, thus providing a monolithic platform for realizing multi-component topological states without requiring separate internal degrees of freedom in the physical setup. Notably, by coupling the defect modulations diagonally, we report the novel realization of an ``entangled'' 4D QH phase which cannot be factorized into two copies of 2D QH phases, each described by the 1st Chern number. The structure of this non-factorizability can be quantified by a classical entanglement entropy inspired by quantum information theory. In another embodiment, we present 4D p-orbital nodal lines in a nonsymmorphic photonic lattice, hosting boundary states with an exotic manifold. Our simple and versatile approach holds the promise of novel topological optoelectronic and photonic applications such as one-way optical fibers.

cond-mat.mes-hall↗

arXiv2019

Graphene/Polyelectrolyte Layer-by-Layer Coatings for Electromagnetic Interference Shielding

Electromagnetic interference (EMI) shielding coating materials with thicknesses in the microscale are required in many sectors, including communications, medical, aerospace and electronics, to isolate the electromagnetic radiation emitted from electronic equipment. We report a spray, layer-by-layer (LbL) coating approach to fabricate micron thick, highly-ordered and electrically-conductive coatings with exceptional EMI shielding effectiveness (EMI SE >4830 dB/mm), through the alternating self-assembly of negatively-charged reduced graphene oxide (RGO) and a positively-charged polyelectrolyte (PEI). The microstructure and resulting electrical properties of the (PEI/RGO)n LbL structures are studied as function of increasing mass of graphene deposited per cycle (keeping the PEI content constant), number of deposited layers (n), flake diameter and type of RGO. A strong effect of the lateral flake dimensions on the electrical properties is observed, which also influences the EMI SE. A maximum EMI SE of 29 dB is obtained for a 6 um thick (PEI/RGO)10 coating with 19 vol.% loading of reduced electrochemically-exfoliated graphene oxide flakes with diameters ~3um. This SE performance exceeds those previously reported for thicker graphene papers and bulk graphene/polymer composite films with higher RGO or graphene nanoplatelets contents, which represents an important step towards the fabrication of thin and light-weight high-performance EMI shielding structures.

physics.app-ph↗

arXiv2019

The iMaterialist Fashion Attribute Dataset

Large-scale image databases such as ImageNet have significantly advanced image classification and other visual recognition tasks. However much of these datasets are constructed only for single-label and coarse object-level classification. For real-world applications, multiple labels and fine-grained categories are often needed, yet very few such datasets exist publicly, especially those of large-scale and high quality. In this work, we contribute to the community a new dataset called iMaterialist Fashion Attribute (iFashion-Attribute) to address this problem in the fashion domain. The dataset is constructed from over one million fashion images with a label space that includes 8 groups of 228 fine-grained attributes in total. Each image is annotated by experts with multiple, high-quality fashion attributes. The result is the first known million-scale multi-label and fine-grained image dataset. We conduct extensive experiments and provide baseline results with modern deep Convolutional Neural Networks (CNNs). Additionally, we demonstrate models pre-trained on iFashion-Attribute achieve superior transfer learning performance on fashion related tasks compared with pre-training from ImageNet or other fashion datasets. Data is available at: https://github.com/visipedia/imat_fashion_comp

cs.CV↗

arXiv2019

Joint Learning of Neural Networks via Iterative Reweighted Least Squares

In this paper, we introduce the problem of jointly learning feed-forward neural networks across a set of relevant but diverse datasets. Compared to learning a separate network from each dataset in isolation, joint learning enables us to extract correlated information across multiple datasets to significantly improve the quality of learned networks. We formulate this problem as joint learning of multiple copies of the same network architecture and enforce the network weights to be shared across these networks. Instead of hand-encoding the shared network layers, we solve an optimization problem to automatically determine how layers should be shared between each pair of datasets. Experimental results show that our approach outperforms baselines without joint learning and those using pretraining-and-fine-tuning. We show the effectiveness of our approach on three tasks: image classification, learning auto-encoders, and image generation.

cs.LG↗

arXiv2019

Delta Embedding Learning

Unsupervised word embeddings have become a popular approach of word representation in NLP tasks. However there are limitations to the semantics represented by unsupervised embeddings, and inadequate fine-tuning of embeddings can lead to suboptimal performance. We propose a novel learning technique called Delta Embedding Learning, which can be applied to general NLP tasks to improve performance by optimized tuning of the word embeddings. A structured regularization is applied to the embeddings to ensure they are tuned in an incremental way. As a result, the tuned word embeddings become better word representations by absorbing semantic information from supervision without "forgetting." We apply the method to various NLP tasks and see a consistent improvement in performance. Evaluation also confirms the tuned word embeddings have better semantic properties.

cs.CL↗

arXiv2019

Sentiment Tagging with Partial Labels using Modular Architectures

Many NLP learning tasks can be decomposed into several distinct sub-tasks, each associated with a partial label. In this paper we focus on a popular class of learning problems, sequence prediction applied to several sentiment analysis tasks, and suggest a modular learning approach in which different sub-tasks are learned using separate functional modules, combined to perform the final task while sharing information. Our experiments show this approach helps constrain the learning process and can alleviate some of the supervision efforts.

cs.CL↗

arXiv2019

AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representations

The cosine-based softmax losses and their variants achieve great success in deep learning based face recognition. However, hyperparameter settings in these losses have significant influences on the optimization path as well as the final recognition performance. Manually tuning those hyperparameters heavily relies on user experience and requires many training tricks. In this paper, we investigate in depth the effects of two important hyperparameters of cosine-based softmax losses, the scale parameter and angular margin parameter, by analyzing how they modulate the predicted classification probability. Based on these analysis, we propose a novel cosine-based softmax loss, AdaCos, which is hyperparameter-free and leverages an adaptive scale parameter to automatically strengthen the training supervisions during the training process. We apply the proposed AdaCos loss to large-scale face verification and identification datasets, including LFW, MegaFace, and IJB-C 1:1 Verification. Our results show that training deep neural networks with the AdaCos loss is stable and able to achieve high face recognition accuracy. Our method outperforms state-of-the-art softmax losses on all the three datasets.

cs.CV↗

arXiv2019

P2SGrad: Refined Gradients for Optimizing Deep Face Models

Cosine-based softmax losses significantly improve the performance of deep face recognition networks. However, these losses always include sensitive hyper-parameters which can make training process unstable, and it is very tricky to set suitable hyper parameters for a specific dataset. This paper addresses this challenge by directly designing the gradients for adaptively training deep neural networks. We first investigate and unify previous cosine softmax losses by analyzing their gradients. This unified view inspires us to propose a novel gradient called P2SGrad (Probability-to-Similarity Gradient), which leverages a cosine similarity instead of classification probability to directly update the testing metrics for updating neural network parameters. P2SGrad is adaptive and hyper-parameter free, which makes the training process more efficient and faster. We evaluate our P2SGrad on three face recognition benchmarks, LFW, MegaFace, and IJB-C. The results show that P2SGrad is stable in training, robust to noise, and achieves state-of-the-art performance on all the three benchmarks.

cs.CV↗

arXiv2019

Electron Acceleration in Middle-age Shell-type $γ$-Ray Supernova Remnants

Over the past decade, $γ$-ray observations of supernova remnants (SNRs) and accurate cosmic-ray (CR) spectral measurements have significantly advanced our understanding of particle acceleration in SNRs. In combination with multiwavelength observations of a large sample of SNRs, it has been proposed that the highest energy particles are mostly accelerated in young remnants, and the maximum energy that middle-age and old SNRs can accelerate particles to decreases rapidly with the decrease in shock speed. If SNRs dominate the CR flux observed at Earth, a large number of particles need to be accelerated in old SNRs for the soft CR spectrum even though they cannot produce very high-energy CRs. With radio, X-ray, and $γ$-ray observations of seven middle-age shell-type SNRs, we derive the distribution of high-energy electrons trapped in these remnants via a simple one-zone leptonic emission model and find that their spectral evolution is consistent with such a scenario. In particular, we find that particle acceleration by shocks in middle-age SNRs with an age of $t$ can be described by a unified model with the maximum energy decreasing as $t^{-3.1}$ and the number of GeV electrons increasing as $t^{2.5}$ in the absence of escape from SNRs.

astro-ph.HE↗

arXiv2019

Fermi LAT detection of the supernova remnant SN 1006 revisited: the south-west limb

The data from the Large Area Telescope (LAT) onboard the Fermi Gamma-ray Space Telescope have recently been updated. We thus re-analyze the LAT data for the supernova remnant (SNR) SN 1006. Two parts of gamma-ray emission from the region is clearly resolved, which correspond to the north-east (NE) and south-west (SW) limbs of the SNR. The former was detected in the previous LAT data (Xing et al. 2016), but the latter is newly detected in this work. The detection of the two limbs are at a 4 sigma significance level, and the spectral results for the NE limb is consistent with those obtained in previous detection analyses. We construct the broadband spectral energy distribution (SED) for the SW limb. Different scenarios are considered for the SED in gamma-ray energies. We conclude very similar to that of the NE limb, the high-energy and very high-energy emission from the SW limb is likely dominated by the leptonic process, in which high-energy electrons accelerated from the shell region of the SNR inverse-Compton scatter background photons to gamma-rays.

astro-ph.HE↗

arXiv2019

On the Vulnerability of CNN Classifiers in EEG-Based BCIs

Deep learning has been successfully used in numerous applications because of its outstanding performance and the ability to avoid manual feature engineering. One such application is electroencephalogram (EEG) based brain-computer interface (BCI), where multiple convolutional neural network (CNN) models have been proposed for EEG classification. However, it has been found that deep learning models can be easily fooled with adversarial examples, which are normal examples with small deliberate perturbations. This paper proposes an unsupervised fast gradient sign method (UFGSM) to attack three popular CNN classifiers in BCIs, and demonstrates its effectiveness. We also verify the transferability of adversarial examples in BCIs, which means we can perform attacks even without knowing the architecture and parameters of the target models, or the datasets they were trained on. To our knowledge, this is the first study on the vulnerability of CNN classifiers in EEG-based BCIs, and hopefully will trigger more attention on the security of BCI systems.

cs.LG↗

arXiv2019

DeeperLab: Single-Shot Image Parser

We present a single-shot, bottom-up approach for whole image parsing. Whole image parsing, also known as Panoptic Segmentation, generalizes the tasks of semantic segmentation for 'stuff' classes and instance segmentation for 'thing' classes, assigning both semantic and instance labels to every pixel in an image. Recent approaches to whole image parsing typically employ separate standalone modules for the constituent semantic and instance segmentation tasks and require multiple passes of inference. Instead, the proposed DeeperLab image parser performs whole image parsing with a significantly simpler, fully convolutional approach that jointly addresses the semantic and instance segmentation tasks in a single-shot manner, resulting in a streamlined system that better lends itself to fast processing. For quantitative evaluation, we use both the instance-based Panoptic Quality (PQ) metric and the proposed region-based Parsing Covering (PC) metric, which better captures the image parsing quality on 'stuff' classes and larger object instances. We report experimental results on the challenging Mapillary Vistas dataset, in which our single model achieves 31.95% (val) / 31.6% PQ (test) and 55.26% PC (val) with 3 frames per second (fps) on GPU or near real-time speed (22.6 fps on GPU) with reduced accuracy.

cs.CV↗

arXiv2019

Is Supernova Remnant Cassiopeia A a PeVatron?

Cassiopeia A, a well-observed young core-collapse supernova remnant (SNR), is considered as one of the best candidates for studying very high-energy particle acceleration up to PeV via the diffusive shock mechanism. Recently, MAGIC observations revealed a $γ$-ray spectral cutoff at $\sim3.5$ TeV, suggesting that if the TeV $γ$-rays have a hadronic origin, SNR Cas A can only accelerate particles to tens of TeV. Here, we propose a two-zone emission model for regions associated with the forward (zone 1) and inward/reverse shocks (zone 2). Given the low density in zone 1, it dominates the high-frequency radio emission, soft X-ray rim via the synchrotron process and TeV $γ$-ray via the inverse Comptonization. With a relatively softer particle distribution and a higher cut-off energy for electrons, emissions from zone 2 dominate the low-frequency radio, hard X-ray via the synchrotron process and GeV $γ$-ray via hadronic processes. There is no evidence for high-energy cutoffs in the proton distributions implying that Cas A can still be a PeVatron. Hadronic processes from zone 1 dominate very high-energy gamma-ray emission. Future observations in hundreds of TeV range can test this model.

astro-ph.HE↗

arXiv2019

Cost-Sensitive Robustness against Adversarial Examples

Several recent works have developed methods for training classifiers that are certifiably robust against norm-bounded adversarial perturbations. These methods assume that all the adversarial transformations are equally important, which is seldom the case in real-world applications. We advocate for cost-sensitive robustness as the criteria for measuring the classifier's performance for tasks where some adversarial transformation are more important than others. We encode the potential harm of each adversarial transformation in a cost matrix, and propose a general objective function to adapt the robust training method of Wong & Kolter (2018) to optimize for cost-sensitive robustness. Our experiments on simple MNIST and CIFAR10 models with a variety of cost matrices show that the proposed approach can produce models with substantially reduced cost-sensitive robust error, while maintaining classification accuracy.

cs.LG↗

arXiv2019

FERMI observation of the jets of the microquasar SS433

We report our analysis of the data obtained with the Large Area Telescope (LAT) onboard the Fermi Gamma-ray Space Telescope (Fermi) for the SS 433/W50 region. The total data length is ten years. Different from the previous results reported for this region (for which an old version of database was used), we show that excess emission is detected around the w1 region in the western lobe of the jets from SS 433. The region is bright at X-rays due to the interaction between the jet and the ambient medium. This detection also matches well the recent results of the very-high-energy detection of SS 433/W50 with the High Altitude Water Cherenkov (HAWC). However, the eastern regions that are slightly brighter in HAWC's observation are not detected in the Fermi data. Constructing the broad-band spectral energy distribution (SED) for the western region, we compare with the HAWC results for the eastern regions and discuss the possible origin of the emission. A leptonic scenario can provide a fit to the Fermi GeV spectrum and HAWC TeV detection, where the former and latter are due to the synchrotron radiation and inverse-Compton process respectively. However, the model can not explain the X-ray and radio emission from the region simultaneously, which thus requires further observational and theoretical studies of the region in order to clarify the reasons.

astro-ph.HE↗

arXiv2019

Helium-hydrogen immiscibility at high pressures

Hydrogen and helium are the most abundant elements in the universe and they constitute the interiors of gas giant planets. Thus, their equations of states, phase, chemical state, and chemical reactivity at extreme conditions are of great interest. Applying Raman spectroscopy, visual observation, and synchrotron X-ray diffraction in diamond anvil cells (DAC), we performed experiments on H2-He 1:1 and D2-He 1:10 compressed gas mixture up to 100 GPa at 300 K. By comparing with the available data on pure bulk materials, we find no sign of miscibility, chemical reactivity, and new compound formation.

cond-mat.mtrl-sci↗

arXiv2019

Spectra of random networks with arbitrary degrees

We derive a message passing method for computing the spectra of locally tree-like networks and an approximation to it that allows us to compute closed-form expressions or fast numerical approximates for the spectral density of random graphs with arbitrary node degrees -- the so-called configuration model. We find the latter approximation to work well for all but the sparsest of networks. We also derive bounds on the position of the band edges of the spectrum, which are important for identifying structural phase transitions in networks.

physics.soc-ph↗

arXiv2018

Topological corner modes in a brick lattice with nonsymmorphic symmetry

The quest for new realizations of higher-order topological system has garnered much recent attention. In this work, we propose a paradigmatic brick lattice model where corner modes requires protection by nonsymmorphic symmetry in addition to two commuting mirror symmetries. Unlike the well-known square corner mode lattice, it has an odd number of occupied bands, which necessitates a different definition for the $\mathbb Z_2\times \mathbb Z_2$ topological invariant. By studying both the quadrupolar polarization and effective edge model, our study culminates in a phase diagram containing two distinct topological regimes. Our brick lattice corner modes can be realized in a RLC circuit setup and detected via collossal "topolectrical" resonances.

cond-mat.str-el↗