arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 1,099 records · Page 61Linked to original sources

arXiv2022

Spatiotemporal Multi-scale Bilateral Motion Network for Gait Recognition

The critical goal of gait recognition is to acquire the inter-frame walking habit representation from the gait sequences. The relations between frames, however, have not received adequate attention in comparison to the intra-frame features. In this paper, motivated by optical flow, the bilateral motion-oriented features are proposed, which can allow the classic convolutional structure to have the capability to directly portray gait movement patterns at the feature level. Based on such features, we develop a set of multi-scale temporal representations that force the motion context to be richly described at various levels of temporal resolution. Furthermore, a correction block is devised to eliminate the segmentation noise of silhouettes for getting more precise gait information. Subsequently, the temporal feature set and the spatial features are combined to comprehensively characterize gait processes. Extensive experiments are conducted on CASIA-B and OU-MVLP datasets, and the results achieve an outstanding identification performance, which has demonstrated the effectiveness of the proposed approach.

cs.CV↗

arXiv2022

Real-Time Power System Event Detection: A Novel Instance Selection Approach

Instance selection is a vital technique for energy big data analytics. It is challenging to process a massive amount of streaming data generated at high speed rates by intelligent monitoring devices. Instance selection aims at removing noisy and bad data that can compromise the performance of data-driven learners. In this context, this paper proposes a novel similarity based instance selection (SIS) method for real-time phasor measurement unit data. In addition, we develop a variant of the Hoeffding-Tree learner enhanced with the SIS for classifying disturbances and cyber-attacks. We validate the merits of the proposed learner by exploring its performance under four scenarios that affect either the system physics or the monitoring architecture. Our experiments are simulated by using the datasets of industrial control system cyber-attacks. Finally, we conduct an implementation analysis which shows the deployment feasibility and high-performance potential of the proposed learner, as a part of real-time monitoring applications.

eess.SP↗

arXiv2022

Discrete Linear Canonical Transform on Graphs

With the wide application of spectral and algebraic theory in discrete signal processing techniques in the field of graph signal processing, an increasing number of signal processing methods have been proposed, such as the graph Fourier transform, graph wavelet transform and windowed graph Fourier transform. In this paper, we propose and design the definition of the discrete linear canonical transform on graphs (GLCT), which is an extension of the discrete linear canonical transform (DLCT), just as the graph Fourier transform (GFT) is an extension of the discrete Fourier transform (DFT). First, based on the centrality and scalability of the DLCT eigendecomposition approach, the definition of GLCT is proposed by combining graph chirp-Fourier transform, graph scale transform and graph fractional Fourier transform. Second, we derive and discuss the properties and special cases of GLCT. Finally, some GLCT examples of the graph signals are given to illustrate the improvement of the transformation.

eess.SP↗

arXiv2022

An Efficient Person Clustering Algorithm for Open Checkout-free Groceries

Open checkout-free grocery is the grocery store where the customers never have to wait in line to check out. Developing a system like this is not trivial since it faces challenges of recognizing the dynamic and massive flow of people. In particular, a clustering method that can efficiently assign each snapshot to the corresponding customer is essential for the system. In order to address the unique challenges in the open checkout-free grocery, we propose an efficient and effective person clustering method. Specifically, we first propose a Crowded Sub-Graph (CSG) to localize the relationship among massive and continuous data streams. CSG is constructed by the proposed Pick-Link-Weight (PLW) strategy, which \textbf{picks} the nodes based on time-space information, \textbf{links} the nodes via trajectory information, and \textbf{weighs} the links by the proposed von Mises-Fisher (vMF) similarity metric. Then, to ensure that the method adapts to the dynamic and unseen person flow, we propose Graph Convolutional Network (GCN) with a simple Nearest Neighbor (NN) strategy to accurately cluster the instances of CSG. GCN is adopted to project the features into low-dimensional separable space, and NN is able to quickly produce a result in this space upon dynamic person flow. The experimental results show that the proposed method outperforms other alternative algorithms in this scenario. In practice, the whole system has been implemented and deployed in several real-world open checkout-free groceries.

cs.CV↗

arXiv2022

Control of chiral orbital currents in a colossal magnetoresistance material

Colossal magnetoresistance (CMR) is an extraordinary enhancement of the electric conductivity in the presence of a magnetic field. It is conventionally associated with a field-induced spin polarization, which drastically reduces spin scattering and thus electric resistance. However, ferrimagnetic Mn3Si2Te6 is an intriguing exception to this rule: it exhibits a 7-order-of-magnitude reduction in ab-plane resistivity with a 13-Tesla anisotropy field which occur only when a magnetic polarization is avoided [1]. Here we report an exotic quantum state that is driven by ab-plane chiral orbital currents (COC) flowing along edges of MnTe6 octahedra. The c-axis orbital moments of ab-plane COC couple to the ferrimagnetic Mn spins to drastically increase the ab-plane conductivity (CMR) when an external magnetic field is aligned along the magnetic hard c-axis. Both the COC state and its CMR are extraordinarily susceptible to small DC currents exceeding a critical threshold, and a hallmark of this COC state is an exotic time-dependent, bistable switching mimicking a first-order melting transition. The control of the COC-enabled CMR and bistable switching offers a fundamentally new paradigm for quantum technologies.

cond-mat.str-el↗

arXiv2022

Semantic Role Labeling as Dependency Parsing: Exploring Latent Tree Structures Inside Arguments

Semantic role labeling (SRL) is a fundamental yet challenging task in the NLP community. Recent works of SRL mainly fall into two lines: 1) BIO-based; 2) span-based. Despite ubiquity, they share some intrinsic drawbacks of not considering internal argument structures, potentially hindering the model's expressiveness. The key challenge is arguments are flat structures, and there are no determined subtree realizations for words inside arguments. To remedy this, in this paper, we propose to regard flat argument spans as latent subtrees, accordingly reducing SRL to a tree parsing task. In particular, we equip our formulation with a novel span-constrained TreeCRF to make tree structures span-aware and further extend it to the second-order case. We conduct extensive experiments on CoNLL05 and CoNLL12 benchmarks. Results reveal that our methods perform favorably better than all previous syntax-agnostic works, achieving new state-of-the-art under both end-to-end and w/ gold predicates settings.

cs.CL↗

arXiv2022

Homotopy pro-nilpotent structured ring spectra and topological Quillen localization

The aim of this paper is to show that homotopy pro-nilpotent structured ring spectra are TQ-local, where structured ring spectra are described as algebras over a spectral operad O. Here, TQ is short for topological Quillen homology, which is weakly equivalent to O-algebra stabilization. An O-algebra is called homotopy pro-nilpotent if it is equivalent to a limit of nilpotent O-algebras. Our result provides new positive evidence to a conjecture by Francis-Gaisgory on Koszul duality for general operads. As an application, we simultaneously extend the previously known 0-connected and nilpotent TQ-Whitehead theorems to a homotopy pro-nilpotent TQ-Whitehead theorem.

math.AT↗

arXiv2022

Fast and Accurate End-to-End Span-based Semantic Role Labeling as Word-based Graph Parsing

This paper proposes to cast end-to-end span-based SRL as a word-based graph parsing task. The major challenge is how to represent spans at the word level. Borrowing ideas from research on Chinese word segmentation and named entity recognition, we propose and compare four different schemata of graph representation, i.e., BES, BE, BIES, and BII, among which we find that the BES schema performs the best. We further gain interesting insights through detailed analysis. Moreover, we propose a simple constrained Viterbi procedure to ensure the legality of the output graph according to the constraints of the SRL structure. We conduct experiments on two widely used benchmark datasets, i.e., CoNLL05 and CoNLL12. Results show that our word-based graph parsing approach achieves consistently better performance than previous results, under all settings of end-to-end and predicate-given, without and with pre-trained language models (PLMs). More importantly, our model can parse 669/252 sentences per second, without and with PLMs respectively.

cs.CL↗

arXiv2022

Colloidal Titanium Nitride Nanobars for Broadband Inexpensive Plasmonics and Photochemistry from Visible to Mid-IR Wavelengths

Developing colloidal plasmonic nanomaterials with high carrier density that show optical resonances and photochemical activity extending from the visible to the mid-infrared (MIR) ranges remains a challenging pursuit. Here, we report the fabrication of titanium nitride (TiN) nanobars obtained using a two step procedure based on a wet chemical route synthesis of TiO2 nanowires and their subsequent high temperature annealing in ammonia flow. Electromagnetic simulations of the resulting TiN nanobars reveal a rich set of optical resonances featuring transverse, longitudinal and mixed transverse longitudinal plasmonic modes that cover energies from the visible to MIR region. TiN nanobars decorated with Pt co catalyst nanocrystals show enhanced photocatalytic hydrogen evolution activity in comparison to both isotropic TiN nanospheres of similar size and TiN nanocubes under near infrared excitation at 940 nm due to the enhanced hot electron generation. We also demonstrate that plasmonic TiN nanobars can be used for the detection of furfural molecular vibrations by providing a strong surface enhanced infrared absorption (SEIRA) effect in the MIR region.

physics.optics↗

arXiv2022

Online Beam Learning with Interference Nulling for Millimeter Wave MIMO Systems

Employing large antenna arrays is a key characteristic of millimeter wave (mmWave) and terahertz communication systems. Due to the hardware constraints and the lack of channel knowledge, codebook based beamforming/combining is normally adopted to achieve the desired array gain. However, most of the existing codebooks focus only on improving the gain of their target user, without taking interference into account. This can incur critical performance degradation in dense networks. In this paper, we propose a sample-efficient online reinforcement learning based beam pattern design algorithm that learns how to shape the beam pattern to null the interfering directions. The proposed approach does not require any explicit channel knowledge or any coordination with the interferers. Simulation results show that the developed solution is capable of learning well-shaped beam patterns that significantly suppress the interference while sacrificing tolerable beamforming/combing gain from the desired user. Furthermore, a hardware proof-of-concept prototype based on mmWave phased arrays is built and used to implement and evaluate the developed online beam learning solutions in realistic scenarios. The learned beam patterns, measured in an anechoic chamber, show the performance gains of the developed framework and highlight a promising machine learning based beam/codebook optimization direction for mmWave and terahertz systems.

eess.SP↗

arXiv2022

Experimental investigation of wave-particle duality relations in asymmetric beam interference

Wave-particle duality relations are fundamental for quantum physics. Previous experimental studies of duality relations mainly focus on the quadratic relation $D^2+V^2\leq1$, based on symmetric beam interference, while a linear form of the duality relation, predicated earlier theoretically, has never been experimentally tested. In addition, the difference between the quadratic form and the linear form has not been explored yet. In this work, with a designed asymmetric beam interference and by utilizing the polarization degree of freedom of the photon as a which-way detector, we experimentally confirm both forms of the duality relations. The results show that more path information is obtained in the quadratic case. Our findings reveal the difference between the two duality relations and have fundamental implications in better understanding these important duality relations.

quant-ph↗

arXiv2022

Quantum state discrimination in a PT-symmetric system

Nonorthogonal quantum state discrimination (QSD) plays an important role in quantum information and quantum communication. In addition, compared to Hermitian quantum systems, parity-time-($\mathcal{PT}$-)symmetric non-Hermitian quantum systems exhibit novel phenomena and have attracted considerable attention. Here, we experimentally demonstrate QSD in a $\mathcal{PT}$-symmetric system (i.e., $\mathcal{PT}$-symmetric QSD), by having quantum states evolve under a $\mathcal{PT}$-symmetric Hamiltonian in a lossy linear optical setup. We observe that two initially nonorthogonal states can rapidly evolve into orthogonal states, and the required evolution time can even be vanishing provided the matrix elements of the Hamiltonian become sufficiently large. We also observe that the cost of such a discrimination is a dissipation of quantum states into the environment. Furthermore, by comparing $\mathcal{PT}$-symmetric QSD with optimal strategies in Hermitian systems, we find that at the critical value, $\mathcal{PT}$-symmetric QSD is equivalent to the optimal unambiguous state discrimination in Hermitian systems. We also extend the $\mathcal{PT}$-symmetric QSD to the case of discriminating three nonorthogonal states. The QSD in a $\mathcal{PT}$-symmetric system opens a new door for quantum state discrimination, which has important applications in quantum computing, quantum cryptography, and quantum communication.

quant-ph↗

arXiv2022

Nonlocal detection of interlayer three-magnon coupling

A leading nonlinear effect in magnonics is the interaction that splits a high-frequency magnon into two low-frequency ones with conserved linear momentum. Here, we report experimental observation of nonlocal three-magnon scattering between spatially separated magnetic systems, viz. a CoFeB nanowire and an yttrium iron garnet (YIG) thin film. Above a certain threshold power of an applied microwave field, a CoFeB Kittel magnon splits into a pair of counter-propagating YIG magnons that induce voltage signals in Pt electrodes on each side, in excellent agreement with model calculations based on the interlayer dipolar interaction. The excited YIG magnon pairs reside mainly in the first excited (n=1) perpdendicular standing spin-wave mode. With increasing power, the n=1 magnons successively scatter into nodeless (n=0) magnons through a four-magnon process. Our results help to assess non-local scattering processes in magnonic circuits that may enable quantum entanglement between distant magnons for quantum information applications.

cond-mat.mes-hall↗

arXiv2022

PFilter: Building Persistent Maps through Feature Filtering for Fast and Accurate LiDAR-based SLAM

Simultaneous localization and mapping (SLAM) based on laser sensors has been widely adopted by mobile robots and autonomous vehicles. These SLAM systems are required to support accurate localization with limited computational resources. In particular, point cloud registration, i.e., the process of matching and aligning multiple LiDAR scans collected at multiple locations in a global coordinate framework, has been deemed as the bottleneck step in SLAM. In this paper, we propose a feature filtering algorithm, PFilter, that can filter out invalid features and can thus greatly alleviate this bottleneck. Meanwhile, the overall registration accuracy is also improved due to the carefully curated feature points. We integrate PFilter into the well-established scan-to-map LiDAR odometry framework, F-LOAM, and evaluate its performance on the KITTI dataset. The experimental results show that PFilter can remove about 48.4% of the points in the local feature map and reduce feature points in scan by 19.3% on average, which save 20.9% processing time per frame. In the mean time, we improve the accuracy by 9.4%.

cs.RO↗

arXiv2022

Training Text-To-Speech Systems From Synthetic Data: A Practical Approach For Accent Transfer Tasks

Transfer tasks in text-to-speech (TTS) synthesis - where one or more aspects of the speech of one set of speakers is transferred to another set of speakers that do not feature these aspects originally - remains a challenging task. One of the challenges is that models that have high-quality transfer capabilities can have issues in stability, making them impractical for user-facing critical tasks. This paper demonstrates that transfer can be obtained by training a robust TTS system on data generated by a less robust TTS system designed for a high-quality transfer task; in particular, a CHiVE-BERT monolingual TTS system is trained on the output of a Tacotron model designed for accent transfer. While some quality loss is inevitable with this approach, experimental results show that the models trained on synthetic data this way can produce high quality audio displaying accent transfer, while preserving speaker characteristics such as speaking style.

cs.SD↗

arXiv2022

Construction of a qudit using Schrodinger cat states and generation of hybrid entanglement between a discrete-variable qudit and a continuous-variable qudit

We show that a continuous-variable (CV) qudit can be constructed using quasiorthogonal cat states of a bosonic mode, when the phase encoded in each cat state is chosen appropriately. With the constructed CV qudit and the discrete-variable (DV) qudit encoded with Fock states, we propose an approach to generate the hybrid maximally entangled state of a CV qudit and a DV qudit by using two microwave cavities coupled to a superconducting flux qutrit. This proposal relies on the initial preparation of a superposition of Fock states of one cavity and the initial preparation of a cat state of the other cavity. After the initial state of each cavity is prepared, this proposal requires only two basic operations, i.e., the first operation employs the dispersive coupling of both cavities with the qutrit while the second operation uses the dispersive coupling of only one cavity with the qutrit. The entangled state production is deterministic and the operation time decreases as the dimensional size of each qudit increases. In addition, during the entire operation, the coupler qutrit remains in the ground state and thus decoherence from the qutrit is significantly reduced. As an example, we further discuss the experimental feasibility for generating the hybrid maximally entangled state of a DV qutrit and a CV qutrit based on circuit QED. This proposal is universal and can be extended to accomplish the same task, by using two microwave or optical cavities coupled to a natural or artificial three-level atom.

quant-ph↗

arXiv2022

Hybrid controlled-SUM gate with one superconducting qutrit and one cat-state qutrit and application in hybrid entangled state preparation

Compared with a qubit, a qudit (i.e., $d$-level or $d$-state quantum system) provides a larger Hilbert space to store and process information. On the other hand, qudit-based hybrid quantum computing usually requires performing hybrid quantum gates with qudits different in their nature or in their encoding format. In this work, we consider the qutrit case, i.e., the case for a qudit with $d$=3. We propose a simple method to realize a hybrid quantum controlled-SUM gate with one superconducting (SC) qutrit and a cat-state qutrit. This gate plus single-qutrit gates form a universal set of ternary logic gates for quantum computing with qutrits. Our proposal is based on circuit QED and operates essentially by employing a SC ququart (a four-level quantum system) dispersively coupled to a microwave cavity. The gate implementation is quite simple because it only requires a single basic operation. Neither classical pulse nor measurement is needed. The auxiliary higher energy level of the SC ququart is virtually excited during the gate operation, thus decoherence from this level is greatly suppressed. As an application of this gate, we discuss the generation of a hybrid maximally-entangled state of one SC qutrit and one cat-state qutrit. We further analyze the experimental feasibility of creating such hybrid entangled state in circuit QED. This proposal is quite general and can be extended to accomplish the same task in a wide range of physical system, such as a four-level natural or artificial atom coupled to an optical or microwave cavity.

quant-ph↗

arXiv2022

Single-step implementation of a hybrid controlled-NOT gate with one superconducting qubit simultaneously controlling multiple target cat-state qubits

Hybrid quantum gates have recently drawn considerable attention. They play significant roles in connecting quantum information processors with qubits of different encoding and have important applications in the transmission of quantum states between a quantum processor and a quantum memory. In this work, we propose a single-step implementation of a multi-target-qubit controlled-NOT gate with one superconducting (SC) qubit simultaneously controlling $n$ target cat-state qubits. In this proposal, the gate is implemented with $n$ microwave cavities coupled to a three-level SC qutrit. The two logic states of the control SC qubit are represented by the two lowest levels of the qutrit, while the two logic states of each target cat-state qubit are represented by two quasi-orthogonal cat states of a microwave cavity. This proposal operates essentially through the dispersive coupling of each cavity with the qutrit. The gate realization is quite simple because it requires only a single-step operation. There is no need of applying a classical pulse or performing a measurement. The gate operation time is independent of the number of target qubits, thus it does not increase as the number of target qubits increases. Moreover, because the third higher energy level of the qutrit is not occupied during the gate operation, decoherence from the qutrit is greatly suppressed. As an application of this hybrid multi-target-qubit gate, we further discuss the generation of a hybrid Greenberger-Horne-Zeilinger (GHZ) entangled state of SC qubits and cat-state qubits. As an example, we numerically analyze the experimental feasibility of generating a hybrid GHZ state of one SC qubit and three cat-state qubits within present circuit QED technology.

quant-ph↗