arXiv subjects

Rong Zhang

Publications and source records attributed to Rong Zhang.

At least 109 records · Page 6Linked to original sources

arXiv2022

Training biases in machine learning for the analytic continuation of quantum many-body Green's functions

We address the problem of analytic continuation of imaginary-frequency Green's functions, which is crucial in many-body physics, using machine learning based on a multi-level residual neural network. We specifically address potential biases that can be introduced due to the use of artificially created spectral functions that are employed to train the neural network. We also implement an uncertainty estimation of the predicted spectral function, based on Monte Carlo dropout, which allows to identify frequency regions where the prediction might not be accurate, and we study the effect of noise, in particular also for situations where the noise level during training is different from that in the actual data. Our analysis demonstrates that this method can indeed achieve a high quality of prediction, comparable or better than the widely used maximum entropy method, but that further improvement is currently limited by the lack of true data that can be used for training. We also benchmark our approach by applying it to the case of SrVO$_3$, where an accurate spectral function has been obtained from dynamical mean-field theory using a solver that works directly on the real frequency axis.

cond-mat.str-el↗

arXiv2022

Synthetic Target Domain Supervision for Open Retrieval QA

Neural passage retrieval is a new and promising approach in open retrieval question answering. In this work, we stress-test the Dense Passage Retriever (DPR) -- a state-of-the-art (SOTA) open domain neural retrieval model -- on closed and specialized target domains such as COVID-19, and find that it lags behind standard BM25 in this important real-world setting. To make DPR more robust under domain shift, we explore its fine-tuning with synthetic training examples, which we generate from unlabeled target domain text using a text-to-text generator. In our experiments, this noisy but fully automated target domain supervision gives DPR a sizable advantage over BM25 in out-of-domain settings, making it a more viable model in practice. Finally, an ensemble of BM25 and our improved DPR model yields the best results, further pushing the SOTA for open retrieval QA on multiple out-of-domain test sets.

cs.CL↗

arXiv2022

Arbitrary coherent distributions in a programmable quantum walk

The coherent superposition of position states in a quantum walk (QW) can be precisely engineered towards the desired distributions to meet the need of quantum information applications. The coherent distribution can make full use of quantum parallel in computation and simulation. Particularly, the uniform superposition provides the robust non-locality, which has wide applications such as the generation of genuine multi-bit random numbers without post-processing. We experimentally demonstrate that the rich dynamics featured with arbitrary coherent distributions can be obtained by introducing different sets of the time- and position-dependent operations. Such a QW is realized by a resource-constant and flexible optical circuit, in which the variable operation is executed based on a Sagnac interferometer in an intrinsically stable and precisely controlled way. Our results contribute to the practical realization of quantum-walk-based quantum computation, quantum simulations and quantum information protocols.

quant-ph↗

arXiv2022

Maximal coin-walker entanglement in a ballistic quantum walk

We report the position-inhomogeneous quantum walk (IQW) can be utilized to produce the maximal high dimensional entanglement while maintaining the quadratic speedup spread of the wave-function. Our calculations show that the maximal coin-walker entanglement can be generated in any odd steps or asymptotically in even steps, and the nearly maximal entanglement can be obtained in even steps after $2$. We implement the IQW by a stable resource-saving time-bin optical network, in which a polarization Sagnac loop is employed to realize the precisely tunable phase shift. Our approach opens up an efficient way for high-dimensional entanglement engineering as well as promotes investigations on the role of coin-walker interactions in QW based applications.

quant-ph↗

arXiv2021

Counterfactual Adversarial Learning with Representation Interpolation

Deep learning models exhibit a preference for statistical fitting over logical reasoning. Spurious correlations might be memorized when there exists statistical bias in training data, which severely limits the model performance especially in small data scenarios. In this work, we introduce Counterfactual Adversarial Training framework (CAT) to tackle the problem from a causality perspective. Particularly, for a specific sample, CAT first generates a counterfactual representation through latent space interpolation in an adversarial manner, and then performs Counterfactual Risk Minimization (CRM) on each original-counterfactual pair to adjust sample-wise loss weight dynamically, which encourages the model to explore the true causal effect. Extensive experiments demonstrate that CAT achieves substantial performance improvement over SOTA across different downstream tasks, including sentence classification, natural language inference and question answering.

cs.LG↗

arXiv2021

Giant Topological Hall Effect in van der Waals Heterostructures of CrTe2/Bi2Te3

Discoveries of interfacial topological Hall effect (THE) provide an ideal platform for exploring physics arising from the interplay between topology and magnetism. The interfacial topological Hall effect is closely related to the Dzyaloshinskii-Moriya interaction (DMI) at interface and topological spin textures. However, it is difficult to achieve a sizable THE in heterostructures due to the stringent constraints on the constituents of THE heterostructures such as strong spin-orbit coupling (SOC). Here we report the observation of a giant THE signal of 1.39 $μΩ\cdot$cm in the van der Waals heterostructures of CrTe2/Bi2Te3 fabricated by molecular beam epitaxy, a prototype of two-dimensional (2D) ferromagnet (FM)/topological insulator (TI). This large magnitude of THE is attributed to an optimized combination of 2D ferromagnetism in CrTe2, strong SOC in Bi2Te3, and an atomically sharp interface. Our work reveals CrTe2/Bi2Te3 as a convenient platform for achieving large interfacial THE in hybrid systems, which could be utilized to develop quantum science and high-density information storage.

cond-mat.mes-hall↗

arXiv2021

High power Figure-of-Merit, 10.6-kV AlGaN/GaN lateral Schottky barrier diode with single channel and sub-100-μm anode-to-cathode spacing

GaN-based lateral Schottky diodes (SBDs) have attracted great attention for high-power applications due to its combined high electron mobility and large critical breakdown field. However, the breakdown voltage (BV) of the SBDs are far from exploiting the material advantages of GaN at present, limiting the desire to use GaN for ultra-high voltage (UHV) applications. Then, a golden question is whether the excellent properties of GaN-based materials can be practically used in the UHV field? Here we demonstrate UHV AlGaN/GaN SBDs on sapphire with a BV of 10.6 kV, a specific on-resistance of 25.8 mΩ.cm2, yielding a power figure of merit of more than 3.8 GW/cm2. These devices are designed with single channel and 85-μm anode-to-cathode spacing, without other additional electric field management, demonstrating its great potential for the UHV application in power electronics.

cond-mat.mtrl-sci↗

arXiv2021

Graph Pattern Loss based Diversified Attention Network for Cross-Modal Retrieval

Cross-modal retrieval aims to enable flexible retrieval experience by combining multimedia data such as image, video, text, and audio. One core of unsupervised approaches is to dig the correlations among different object representations to complete satisfied retrieval performance without requiring expensive labels. In this paper, we propose a Graph Pattern Loss based Diversified Attention Network(GPLDAN) for unsupervised cross-modal retrieval to deeply analyze correlations among representations. First, we propose a diversified attention feature projector by considering the interaction between different representations to generate multiple representations of an instance. Then, we design a novel graph pattern loss to explore the correlations among different representations, in this graph all possible distances between different representations are considered. In addition, a modality classifier is added to explicitly declare the corresponding modalities of features before fusion and guide the network to enhance discrimination ability. We test GPLDAN on four public datasets. Compared with the state-of-the-art cross-modal retrieval methods, the experimental results demonstrate the performance and competitiveness of GPLDAN.

cs.CV↗

arXiv2021

Gate control of spin-layer-locking FETs and application to monolayer LuIO

A recent 2D spinFET concept proposes to switch electrostatically between two separate sublayers with strong and opposite intrinsic Rashba effects. This concept exploits the spin-layer locking mechanism present in centrosymmetric materials with local dipole fields, where a weak electric field can easily manipulate just one of the spin channels. Here, we propose a novel monolayer material within this family, lutetium oxide iodide (LuIO). It displays one of the largest Rashba effects among 2D materials (up to $k_R = 0.08$ Å$^{-1}$), leading to a $π/2$ rotation of the spins over just 1 nm. The monolayer had been predicted to be exfoliable from its experimentally-known 3D bulk counterpart, with a binding energy even lower than graphene. We characterize and model with first-principles simulations the interplay of the two gate-controlled parameters for such devices: doping and spin channel selection. We show that the ability to split the spin channels in energy diminishes with doping, leading to specific gate-operation guidelines that can apply to all devices based on spin-layer locking.

cond-mat.mtrl-sci↗

arXiv2021

Localized surface plasmon enhanced InGaN-based light-emitting diodes with polygonal Au/Al nanoparticles

Localized surface plasmons (LSPs) have played a significant role in improving the light emission efficiency of light emitting diodes (LEDs). In this report, polygonal nanoholes have been fabricated in the p-GaN layer of InGaN-based LEDs by using Ni nanoporous film as the etching mask, and then Au/Al metal nanoparticles are embedded in the nanoholes to form the LSP structure. The coupling between the LSP and the LED has been clearly observed. The results show that the light output of the LEDs has been increased by 46% at higher current injection condition, and together with a shift of the gain peak position to the LSP peak resonance energy. As the coupling distance is decreased from 60 nm to 30 nm, the maximum enhancement factor increases to 2.38. The above results indicate that the LSP from the polygonal metal nanoparticles is a kind of very promising structure to enhance the lighting performance of the InGaN-based LEDs.

physics.optics↗

arXiv2021

Enhancing Model Robustness By Incorporating Adversarial Knowledge Into Semantic Representation

Despite that deep neural networks (DNNs) have achieved enormous success in many domains like natural language processing (NLP), they have also been proven to be vulnerable to maliciously generated adversarial examples. Such inherent vulnerability has threatened various real-world deployed DNNs-based applications. To strength the model robustness, several countermeasures have been proposed in the English NLP domain and obtained satisfactory performance. However, due to the unique language properties of Chinese, it is not trivial to extend existing defenses to the Chinese domain. Therefore, we propose AdvGraph, a novel defense which enhances the robustness of Chinese-based NLP models by incorporating adversarial knowledge into the semantic representation of the input. Extensive experiments on two real-world tasks show that AdvGraph exhibits better performance compared with previous work: (i) effective - it significantly strengthens the model robustness even under the adaptive attacks setting without negative impact on model performance over legitimate input; (ii) generic - its key component, i.e., the representation of connotative adversarial knowledge is task-agnostic, which can be reused in any Chinese-based NLP models without retraining; and (iii) efficient - it is a light-weight defense with sub-linear computational complexity, which can guarantee the efficiency required in practical scenarios.

cs.CL↗

arXiv2020

The Open Brands Dataset: Unified brand detection and recognition at scale

Intellectual property protection(IPP) have received more and more attention recently due to the development of the global e-commerce platforms. brand recognition plays a significant role in IPP. Recent studies for brand recognition and detection are based on small-scale datasets that are not comprehensive enough when exploring emerging deep learning techniques. Moreover, it is challenging to evaluate the true performance of brand detection methods in realistic and open scenes. In order to tackle these problems, we first define the special issues of brand detection and recognition compared with generic object detection. Second, a novel brands benchmark called "Open Brands" is established. The dataset contains 1,437,812 images which have brands and 50,000 images without any brand. The part with brands in Open Brands contains 3,113,828 instances annotated in 3 dimensions: 4 types, 559 brands and 1216 logos. To the best of our knowledge, it is the largest dataset for brand detection and recognition with rich annotations. We provide in-depth comprehensive statistics about the dataset, validate the quality of the annotations and study how the performance of many modern models evolves with an increasing amount of training data. Third, we design a network called "Brand Net" to handle brand recognition. Brand Net gets state-of-art mAP on Open Brand compared with existing detection methods.

cs.CV↗

arXiv2020

The Fictitious Domain Method Based on Navier Slip Boundary Condition for Simulation of Flow-Particle Interaction

In this article, we develop a least--squares/fictitious domain method for direct simulation of fluid particle motion with Navier slip boundary condition at the fluid--particle interface. Let $Ω$ and $B$ be two bounded domains of $\mathbb{R}^{d}$ such that $\overline{B} \subset Ω$. The motion of solid particle $B$ is governed by Newton's equations. Our goal here is to develop a fictitious domain method where one solves a variant of the original problem on the full $Ω$, followed by a well--chosen correction over $B$ and corrections related to translation velocity and angular velocity of the particle. This method is of the virtual control type and relies on a least--squares formulation making the problem solvable by a conjugate gradient algorithm operating in a well chosen control space. Since the fully explicit scheme to update the particle motion using Newton's equation is unstable, we propose and implement an explicit--implicit scheme in which, at each time step, the position of the particle is updated explicitly, and the solution of Navier-Stokes equations and particle velocities are solved by the the least--squares/fictitious domain method implicitly. Numerical results are given to verify our numerical method.

math.NA↗

arXiv2020

End-to-End QA on COVID-19: Domain Adaptation with Synthetic Training

End-to-end question answering (QA) requires both information retrieval (IR) over a large document collection and machine reading comprehension (MRC) on the retrieved passages. Recent work has successfully trained neural IR systems using only supervised question answering (QA) examples from open-domain datasets. However, despite impressive performance on Wikipedia, neural IR lags behind traditional term matching approaches such as BM25 in more specific and specialized target domains such as COVID-19. Furthermore, given little or no labeled data, effective adaptation of QA systems can also be challenging in such target domains. In this work, we explore the application of synthetically generated QA examples to improve performance on closed-domain retrieval and MRC. We combine our neural IR and MRC systems and show significant improvements in end-to-end QA on the CORD-19 collection over a state-of-the-art open-domain QA baseline.

cs.CL↗

arXiv2020

Answer Span Correction in Machine Reading Comprehension

Answer validation in machine reading comprehension (MRC) consists of verifying an extracted answer against an input context and question pair. Previous work has looked at re-assessing the "answerability" of the question given the extracted answer. Here we address a different problem: the tendency of existing MRC systems to produce partially correct answers when presented with answerable questions. We explore the nature of such errors and propose a post-processing correction method that yields statistically significant performance improvements over state-of-the-art MRC systems in both monolingual and multilingual evaluation.

cs.CL↗

arXiv2020

AADS: Augmented Autonomous Driving Simulation using Data-driven Algorithms

Simulation systems have become an essential component in the development and validation of autonomous driving technologies. The prevailing state-of-the-art approach for simulation is to use game engines or high-fidelity computer graphics (CG) models to create driving scenarios. However, creating CG models and vehicle movements (e.g., the assets for simulation) remains a manual task that can be costly and time-consuming. In addition, the fidelity of CG images still lacks the richness and authenticity of real-world images and using these images for training leads to degraded performance. In this paper we present a novel approach to address these issues: Augmented Autonomous Driving Simulation (AADS). Our formulation augments real-world pictures with a simulated traffic flow to create photo-realistic simulation images and renderings. More specifically, we use LiDAR and cameras to scan street scenes. From the acquired trajectory data, we generate highly plausible traffic flows for cars and pedestrians and compose them into the background. The composite images can be re-synthesized with different viewpoints and sensor models. The resulting images are photo-realistic, fully annotated, and ready for end-to-end training and testing of autonomous driving systems from perception to planning. We explain our system design and validate our algorithms with a number of autonomous driving tasks from detection to segmentation and predictions. Compared to traditional approaches, our method offers unmatched scalability and realism. Scalability is particularly important for AD simulation and we believe the complexity and diversity of the real world cannot be realistically captured in a virtual environment. Our augmented approach combines the flexibility in a virtual environment (e.g., vehicle movements) with the richness of the real world to allow effective simulation of anywhere on earth.

cs.CV↗

arXiv2020

Multi-Stage Pre-training for Low-Resource Domain Adaptation

Transfer learning techniques are particularly useful in NLP tasks where a sizable amount of high-quality annotated data is difficult to obtain. Current approaches directly adapt a pre-trained language model (LM) on in-domain text before fine-tuning to downstream tasks. We show that extending the vocabulary of the LM with domain-specific terms leads to further gains. To a bigger effect, we utilize structure in the unlabeled data to create auxiliary synthetic tasks, which helps the LM transfer to downstream tasks. We apply these approaches incrementally on a pre-trained Roberta-large LM and show considerable performance gain on three tasks in the IT domain: Extractive Reading Comprehension, Document Ranking and Duplicate Question Detection.

cs.CL↗

arXiv2020

TaxThemis: Interactive Mining and Exploration of Suspicious Tax Evasion Group

Tax evasion is a serious economic problem for many countries, as it can undermine the government' s tax system and lead to an unfair business competition environment. Recent research has applied data analytics techniques to analyze and detect tax evasion behaviors of individual taxpayers. However, they failed to support the analysis and exploration of the uprising related party transaction tax evasion (RPTTE) behaviors (e.g., transfer pricing), where a group of taxpayers is involved. In this paper, we present TaxThemis, an interactive visual analytics system to help tax officers mine and explore suspicious tax evasion groups through analyzing heterogeneous tax-related data. A taxpayer network is constructed and fused with the trade network to detect suspicious RPTTE groups. Rich visualizations are designed to facilitate the exploration and investigation of suspicious transactions between related taxpayers with profit and topological data analysis. Specifically, we propose a calendar heatmap with a carefully-designed encoding scheme to intuitively show the evidence of transferring revenue through related party transactions. We demonstrate the usefulness and effectiveness of TaxThemis through two case studies on real-world tax-related data, and interviews with domain experts.

cs.HC↗