arXiv subjects

Rui Zhang

Publications and source records attributed to Rui Zhang.

At least 1,027 records · Page 57Linked to original sources

arXiv2020

A heterogeneous branch and multi-level classification network for person re-identification

Convolutional neural networks with multiple branches have recently been proved highly effective in person re-identification (re-ID). Researchers design multi-branch networks using part models, yet they always attribute the effectiveness to multiple parts. In addition, existing multi-branch networks always have isomorphic branches, which lack structural diversity. In order to improve this problem, we propose a novel Heterogeneous Branch and Multi-level Classification Network (HBMCN), which is designed based on the pre-trained ResNet-50 model. A new heterogeneous branch, SE-Res-Branch, is proposed based on the SE-Res module, which consists of the Squeeze-and-Excitation block and the residual block. Furthermore, a new multi-level classification joint objective function is proposed for the supervised learning of HBMCN, whereby multi-level features are extracted from multiple high-level layers and concatenated to represent a person. Based on three public person re-ID benchmarks (Market1501, DukeMTMC-reID and CUHK03), experimental results show that the proposed HBMCN reaches 94.4%, 85.7% and 73.8% in Rank-1, and 85.7%, 74.6% and 69.0% in mAP, achieving a state-of-the-art performance. Further analysis demonstrates that the specially designed heterogeneous branch performs better than an isomorphic branch, and multi-level classification provides more discriminative features compared to single-level classification. As a result, HBMCN provides substantial further improvements in person re-ID tasks.

cs.CV↗

arXiv2020

Uplink Cooperative Interference Cancellation for Cellular-Connected UAV: A Quantize-and-Forward Approach

Aerial-ground interference is the main obstacle to achieving high spectral efficiency in cellular networks with both traditional terrestrial users and new unmanned aerial vehicle (UAV) users. Due to their strong line-of-sight (LoS) channels with the ground, UAVs could cause/suffer severe interference to/from a large number of non-associated but co-channel base stations (BSs) in their uplink/downlink communications. In this letter, we propose a new cooperative interference cancellation (CIC) scheme for the UAV's uplink communication to mitigate its strong interference to co-channel BSs, which requires only local cooperation between each co-channel BS and its adjacent helping BSs. Specifically, the helping BSs without serving any users in the UAV's communication channel quantize and forward their received signals from the UAV to their aided co-channel BS, which then processes the quantized signals jointly with its own received signal to decode its served terrestrial user's message via canceling the UAV's signal by linear/nonlinear interference cancellation techniques (ICTs). We derive the achievable rates of the proposed CIC scheme with different ICTs as a function of the UAV's transmit power and rate, and thereby unveil the conditions under which the proposed CIC scheme outperforms the existing CIC scheme based on the decode-and-forward (DF) operation of the helping BSs.

cs.IT↗

arXiv2020

BrePartition: Optimized High-Dimensional kNN Search with Bregman Distances

Bregman distances (also known as Bregman divergences) are widely used in machine learning, speech recognition and signal processing, and kNN searches with Bregman distances have become increasingly important with the rapid advances of multimedia applications. Data in multimedia applications such as images and videos are commonly transformed into space of hundreds of dimensions. Such high-dimensional space has posed significant challenges for existing kNN search algorithms with Bregman distances, which could only handle data of medium dimensionality (typically less than 100). This paper addresses the urgent problem of high-dimensional kNN search with Bregman distances. We propose a novel partition-filter-refinement framework. Specifically, we propose an optimized dimensionality partitioning scheme to solve several non-trivial issues. First, an effective bound from each partitioned subspace to obtain exact kNN results is derived. Second, we conduct an in-depth analysis of the optimized number of partitions and devise an effective strategy for partitioning. Third, we design an efficient integrated index structure for all the subspaces together to accelerate the search processing. Moreover, we extend our exact solution to an approximate version by a trade-off between the accuracy and efficiency. Experimental results on four real-world datasets and two synthetic datasets show the clear advantage of our method in comparison to state-of-the-art algorithms.

cs.DB↗

arXiv2020

Pion and Kaon Distribution Amplitudes in the Continuum Limit

We present a lattice-QCD calculation of the pion, kaon and $η_s$ distribution amplitudes using large-momentum effective theory (LaMET). Our calculation is carried out using three ensembles with 2+1+1 flavors of highly improved staggered quarks (HISQ), generated by MILC collaboration, at 310 MeV pion mass with 0.06, 0.09 and 0.12 fm lattice spacings. We use clover fermion action for the valence quarks and tune the quark mass to match the lightest light and strange masses in the sea. The resulting lattice matrix elements are nonperturbatively renormalized in regularization-independent momentum-subtraction (RI/MOM) scheme and extrapolated to the continuum. We use two approaches to extract the $x$-dependence of the meson distribution amplitudes: 1) we fit the renormalized matrix elements in coordinate space to an assumed distribution form through a one-loop matching kernel; 2) we use a machine-learning algorithm trained on pseudo lattice-QCD data to make predictions on the lattice data. We found the results are consistent between these methods with the latter method giving a less smooth shape. Both approaches suggest that as the quark mass increases, the distribution amplitude becomes narrower. Our pion distribution amplitude has broader distribution than predicted by light-front constituent-quark model, and the moments of our pion distributions agree with previous lattice-QCD results using the operator production expansion.

hep-lat↗

arXiv2020

Map Generation from Large Scale Incomplete and Inaccurate Data Labels

Accurately and globally mapping human infrastructure is an important and challenging task with applications in routing, regulation compliance monitoring, and natural disaster response management etc.. In this paper we present progress in developing an algorithmic pipeline and distributed compute system that automates the process of map creation using high resolution aerial images. Unlike previous studies, most of which use datasets that are available only in a few cities across the world, we utilizes publicly available imagery and map data, both of which cover the contiguous United States (CONUS). We approach the technical challenge of inaccurate and incomplete training data adopting state-of-the-art convolutional neural network architectures such as the U-Net and the CycleGAN to incrementally generate maps with increasingly more accurate and more complete labels of man-made infrastructure such as roads and houses. Since scaling the mapping task to CONUS calls for parallelization, we then adopted an asynchronous distributed stochastic parallel gradient descent training scheme to distribute the computational workload onto a cluster of GPUs with nearly linear speed-up.

cs.CV↗

arXiv2020

ESPRIT: Explaining Solutions to Physical Reasoning Tasks

Neural networks lack the ability to reason about qualitative physics and so cannot generalize to scenarios and tasks unseen during training. We propose ESPRIT, a framework for commonsense reasoning about qualitative physics in natural language that generates interpretable descriptions of physical events. We use a two-step approach of first identifying the pivotal physical events in an environment and then generating natural language descriptions of those events using a data-to-text approach. Our framework learns to generate explanations of how the physical simulation will causally evolve so that an agent or a human can easily reason about a solution using those interpretable descriptions. Human evaluations indicate that ESPRIT produces crucial fine-grained details and has high coverage of physical concepts compared to even human annotations. Dataset, code and documentation are available at https://github.com/salesforce/esprit.

cs.CL↗

arXiv2020

Counting Classical Nodes in Quantum Networks

Quantum networks illustrate the use of connected nodes of quantum systems as the backbone of distributed quantum information processing. When the network nodes are entangled in graph states, such a quantum platform is indispensable to almost all the existing distributed quantum tasks. Unfortunately, real networks unavoidably suffer from noise and technical restrictions, making nodes transit from quantum to classical at worst. Here, we introduce a figure of merit in terms of the number of classical nodes for quantum networks in arbitrary graph states. Such a network property is revealed by exploiting a novel Einstein-Podolsky-Rosen steerability. Experimentally, we demonstrate photonic quantum networks of $n_q$ quantum nodes and $n_c$ classical nodes with $n_q$ up to 6 and $n_c$ up to 18 using spontaneous parametric down-conversion entanglement sources. We show that the proposed method is faithful in quantifying the classical defects in prepared multiphoton quantum networks. Our results provide novel identification of generic quantum networks and nonclassical correlations in graph states.

quant-ph↗

arXiv2020

Semi-Supervised Dialogue Policy Learning via Stochastic Reward Estimation

Dialogue policy optimization often obtains feedback until task completion in task-oriented dialogue systems. This is insufficient for training intermediate dialogue turns since supervision signals (or rewards) are only provided at the end of dialogues. To address this issue, reward learning has been introduced to learn from state-action pairs of an optimal policy to provide turn-by-turn rewards. This approach requires complete state-action annotations of human-to-human dialogues (i.e., expert demonstrations), which is labor intensive. To overcome this limitation, we propose a novel reward learning approach for semi-supervised policy learning. The proposed approach learns a dynamics model as the reward function which models dialogue progress (i.e., state-action sequences) based on expert demonstrations, either with or without annotations. The dynamics model computes rewards by predicting whether the dialogue progress is consistent with expert demonstrations. We further propose to learn action embeddings for a better generalization of the reward function. The proposed approach outperforms competitive policy learning baselines on MultiWOZ, a benchmark multi-domain dataset.

cs.CL↗

arXiv2020

Probing nucleon strange and charm distributions with lattice QCD

We present the first lattice-QCD calculation of the unpolarized strange and charm parton distribution functions using large-momentum effective theory (LaMET). We use a lattice ensemble with 2+1+1 flavors of highly improved staggered quarks (HISQ) generated by MILC collaboration, with lattice spacing $a\approx 0.12$ fm and $M_π\approx 310$ MeV, and clover valence fermions with two valence pion masses: 310 and 690 MeV. We use momentum-smeared sources to improve the signal up to nucleon boost momentum $P_z=2.18$ GeV, and determine nonperturbative renormalization factors in RI/MOM scheme. We compare our lattice results with the matrix elements obtained from matching the PDFs from CT18NNLO and NNPDF3.1NNLO global fits. Our data support the assumptions of strange-antistrange and charm-anticharm symmetry that are commonly used in global PDF fits, and we find smaller than expected parton distribution at mid to small $x$.

hep-lat↗

arXiv2020

Security of Distributed Machine Learning: A Game-Theoretic Approach to Design Secure DSVM

Distributed machine learning algorithms play a significant role in processing massive data sets over large networks. However, the increasing reliance on machine learning on information and communication technologies (ICTs) makes it inherently vulnerable to cyber threats. This work aims to develop secure distributed algorithms to protect the learning from data poisoning and network attacks. We establish a game-theoretic framework to capture the conflicting goals of a learner who uses distributed support vector machines (SVMs) and an attacker who is capable of modifying training data and labels. We develop a fully distributed and iterative algorithm to capture real-time reactions of the learner at each node to adversarial behaviors. The numerical results show that distributed SVM is prone to fail in different types of attacks, and their impact has a strong dependence on the network structure and attack capabilities.

cs.CR↗

arXiv2020

Holographic MIMO Surfaces for 6G Wireless Networks: Opportunities, Challenges, and Trends

Future wireless networks are expected to evolve towards an intelligent and software reconfigurable paradigm enabling ubiquitous communications between humans and mobile devices. They will be also capable of sensing, controlling, and optimizing the wireless environment to fulfill the visions of low-power, high-throughput, massively-connected, and low-latency communications. A key conceptual enabler that is recently gaining increasing popularity is the Holographic Multiple Input Multiple Output Surface (HMIMOS) that refers to a low-cost transformative wireless planar structure comprising of sub-wavelength metallic or dielectric scattering particles, which is capable of impacting electromagnetic waves according to desired objectives. In this article, we provide an overview of HMIMOS communications by introducing the available hardware architectures for reconfigurable such metasurfaces and their main characteristics, as well as highlighting the opportunities and key challenges in designing HMIMOS-enabled communications.

cs.IT↗

arXiv2020

Portable Intrinsic Gradiometer for Ultra-Sensitive Detection of Magnetic Gradient in Unshielded Environment

We demonstrate a portable all-optical intrinsic scalar magnetic gradiometer composed of miniaturized cesium vapor cells and vertical-cavity surface-emitting lasers (VCSELs). Two cells, with an inner dimension of 5 mm x 5 mm x 5 mm and separated by a baseline of 5 cm, are driven by one VCSEL and the resulting Larmor precessions are probed by a second VCSEL through optical rotation. The off-resonant linearly polarized probe light interrogates two cells at the same time and the output of the intrinsic gradiometer is proportional to the magnetic field gradient measured over the given baseline. This intrinsic gradiometer scheme has the advantage of avoiding added noise from combining two scalar magnetometers. We achieve better than 18 fT/cm/rt-Hz sensitivity in the gradient measurement. Ultra-sensitive short-baseline magnetic gradiometers can potentially play an important role in many practical applications, such as nondestructive evaluation and unexploded ordnance (UXO) detection. Another application of the gradiometer is for magnetocardiography (MCG) in an unshielded environment. Real-time MCG signals can be extracted from the raw gradiometer readings. The demonstrated gradiometer greatly simplifies the MCG setup and may lead to ubiquitous MCG measurement in the future.

physics.atom-ph↗

arXiv2020

ReADS: A Rectified Attentional Double Supervised Network for Scene Text Recognition

In recent years, scene text recognition is always regarded as a sequence-to-sequence problem. Connectionist Temporal Classification (CTC) and Attentional sequence recognition (Attn) are two very prevailing approaches to tackle this problem while they may fail in some scenarios respectively. CTC concentrates more on every individual character but is weak in text semantic dependency modeling. Attn based methods have better context semantic modeling ability while tends to overfit on limited training data. In this paper, we elaborately design a Rectified Attentional Double Supervised Network (ReADS) for general scene text recognition. To overcome the weakness of CTC and Attn, both of them are applied in our method but with different modules in two supervised branches which can make a complementary to each other. Moreover, effective spatial and channel attention mechanisms are introduced to eliminate background noise and extract valid foreground information. Finally, a simple rectified network is implemented to rectify irregular text. The ReADS can be trained end-to-end and only word-level annotations are required. Extensive experiments on various benchmarks verify the effectiveness of ReADS which achieves state-of-the-art performance.

cs.CV↗

arXiv2020

Cooperative Double-IRS Aided Communication: Beamforming Design and Power Scaling

Intelligent reflecting surface (IRS) is a promising technology to support high performance wireless communication. By adaptively configuring the reflection amplitude and/or phase of each passive reflecting element on it, the IRS can reshape the electromagnetic environment in favour of signal transmission. This letter advances the existing research by proposing and analyzing a double-IRS aided wireless communication system. Under the reasonable assumption that the reflection channel from IRS 1 to IRS 2 is of rank 1 (e.g., line-of-sight channel), we propose a joint passive beamforming design for the two IRSs. Based on this, we show that deploying two cooperative IRSs with in total K elements can yield a power gain of order O(K^4), which greatly outperforms the case of deploying one traditional IRS with a power gain of order O(K^2). Our simulation results validate that the performance of deploying two cooperative IRSs is significantly better than that of deploying one IRS given a sufficient total number of IRS elements. We also extend our line-of-sight channel model to show how different channel models affect the performance of the double-IRS aided wireless communication system.

cs.IT↗

arXiv2020

Channel Estimation and Passive Beamforming for Intelligent Reflecting Surface: Discrete Phase Shift and Progressive Refinement

Prior studies on Intelligent Reflecting Surface (IRS) have mostly assumed perfect channel state information (CSI) available for designing the IRS passive beamforming as well as the continuously adjustable phase shift at each of its reflecting elements, which, however, have simplified two challenging issues for implementing IRS in practice, namely, its channel estimation and passive beamforming designs both under the constraint of discrete phase shifts. To address them, we consider in this paper an IRS-aided single-user communication system with discrete phase shifts and design the IRS training reflection matrix for channel estimation as well as the passive beamforming for data transmission, both subject to the constraint of discrete phase shifts. We show that the training reflection matrix design for discrete phase shifts greatly differs from that for continuous phase shifts, and thus the corresponding passive beamforming should be optimized by taking into account the correlated channel estimation error due to discrete phase shifts. Specifically, we consider a practical block-based transmission, where each block has a finite (insufficient) number of training symbols for channel estimation. A novel hierarchical training reflection design is proposed to progressively estimate IRS elements' channels over multiple blocks by exploiting IRS-elements grouping and partition. Based on the resolved IRS channels in each block, we further design the progressive passive beamforming at the IRS with discrete phase shifts to improve the achievable rate for data transmission over the blocks.

cs.IT↗

arXiv2020

Creation of 2000-atom Greenberger-Horne-Zeilinger states by entanglement amplification

We propose a novel entanglement-creation scheme in a multi-atom ensemble, named entanglement amplification, which converts unentangled states into entangled states and amplifies less-entangled ones to maximally-entangled Greenberger-Horne-Zeilinger (GHZ) states. The scheme starts with a multi-atom ensemble initialized in a coherent spin state. By shifting the energy of a particular Dicke state, we break the Hilbert space of the ensemble into two isolated subspaces to tear the coherent spin state into two components so that entanglement is introduced. After that, we utilize the isolated subspaces to further enhance the entanglement by coherently separating the two components. By single-particle Rabi drivings on atoms in a high-finesse optical cavity illuminated by a single-frequency light, 2000-atom GHZ states can be created with a fidelity above 80% in an experimentally achievable system, making resources of ensembles at Heisenberg limit practically available for quantum metrology.

quant-ph↗

arXiv2020

RF-Rhythm: Secure and Usable Two-Factor RFID Authentication

Passive RFID technology is widely used in user authentication and access control. We propose RF-Rhythm, a secure and usable two-factor RFID authentication system with strong resilience to lost/stolen/cloned RFID cards. In RF-Rhythm, each legitimate user performs a sequence of taps on his/her RFID card according to a self-chosen secret melody. Such rhythmic taps can induce phase changes in the backscattered signals, which the RFID reader can detect to recover the user's tapping rhythm. In addition to verifying the RFID card's identification information as usual, the backend server compares the extracted tapping rhythm with what it acquires in the user enrollment phase. The user passes authentication checks if and only if both verifications succeed. We also propose a novel phase-hopping protocol in which the RFID reader emits Continuous Wave (CW) with random phases for extracting the user's secret tapping rhythm. Our protocol can prevent a capable adversary from extracting and then replaying a legitimate tapping rhythm from sniffed RFID signals. Comprehensive user experiments confirm the high security and usability of RF-Rhythm with false-positive and false-negative rates close to zero.

eess.SP↗

arXiv2020

Simultaneous Navigation and Radio Mapping for Cellular-Connected UAV with Deep Reinforcement Learning

Cellular-connected unmanned aerial vehicle (UAV) is a promising technology to unlock the full potential of UAVs in the future. However, how to achieve ubiquitous three-dimensional (3D) communication coverage for the UAVs in the sky is a new challenge. In this paper, we tackle this challenge by a new coverage-aware navigation approach, which exploits the UAV's controllable mobility to design its navigation/trajectory to avoid the cellular BSs' coverage holes while accomplishing their missions. We formulate an UAV trajectory optimization problem to minimize the weighted sum of its mission completion time and expected communication outage duration, and propose a new solution approach based on the technique of deep reinforcement learning (DRL). To further improve the performance, we propose a new framework called simultaneous navigation and radio mapping (SNARM), where the UAV's signal measurement is used not only for training the deep Q network (DQN) directly, but also to create a radio map that is able to predict the outage probabilities at all locations in the area of interest. This thus enables the generation of simulated UAV trajectories and predicting their expected returns, which are then used to further train the DQN via Dyna technique, thus greatly improving the learning efficiency.

eess.SP↗