arXiv subjects

Yi Chen

Publications and source records attributed to Yi Chen.

At least 253 records · Page 14Linked to original sources

arXiv2021

Harnessing the Quantum Behavior of Spins on Surfaces

The desire to control and measure individual quantum systems such as atoms and ions in a vacuum has led to significant scientific and engineering developments in the past decades that form the basis of today's quantum information science. Single atoms and molecules on surfaces, on the other hand, are heavily investigated by physicists, chemists, and material scientists in search of novel electronic and magnetic functionalities. These two paths crossed in 2015 when it was first clearly demonstrated that individual spins on a surface can be coherently controlled and read out in an all-electrical fashion. The enabling technique is a combination of scanning tunneling microscopy (STM) and electron spin resonance (ESR), which offers unprecedented coherent controllability at the Angstrom length scale. This review aims to illustrate the essential ingredients that allow the quantum operations of single spins on surfaces. Three domains of applications of surface spins, namely quantum sensing, quantum control, and quantum simulation, are discussed with physical principles explained and examples presented. Enabled by the atomically-precise fabrication capability of STM, single spins on surfaces might one day lead to the realization of quantum nanodevices and artificial quantum materials at the atomic scale.

physics.app-ph↗

arXiv2021

Improved YOLOv5 network for real-time multi-scale traffic sign detection

Traffic sign detection is a challenging task for the unmanned driving system, especially for the detection of multi-scale targets and the real-time problem of detection. In the traffic sign detection process, the scale of the targets changes greatly, which will have a certain impact on the detection accuracy. Feature pyramid is widely used to solve this problem but it might break the feature consistency across different scales of traffic signs. Moreover, in practical application, it is difficult for common methods to improve the detection accuracy of multi-scale traffic signs while ensuring real-time detection. In this paper, we propose an improved feature pyramid model, named AF-FPN, which utilizes the adaptive attention module (AAM) and feature enhancement module (FEM) to reduce the information loss in the process of feature map generation and enhance the representation ability of the feature pyramid. We replaced the original feature pyramid network in YOLOv5 with AF-FPN, which improves the detection performance for multi-scale targets of the YOLOv5 network under the premise of ensuring real-time detection. Furthermore, a new automatic learning data augmentation method is proposed to enrich the dataset and improve the robustness of the model to make it more suitable for practical scenarios. Extensive experimental results on the Tsinghua-Tencent 100K (TT100K) dataset demonstrate the effectiveness and superiority of the proposed method when compared with several state-of-the-art methods.

cs.CV↗

arXiv2021

A Framework of Mahalanobis-Distance Metric with Supervised Learning for Clustering Multipath Components in MIMO Channel Analysis

As multipath components (MPCs) are experimentally observed to appear in clusters, cluster-based channel models have been focused in the wireless channel study. However, most of the MPC clustering algorithms for MIMO channels with delay and angle information of MPCs are based on the distance metric that quantifies the similarity of two MPCs and determines the preferred cluster shape, greatly impacting MPC clustering quality. In this paper, a general framework of Mahalanobis-distance metric is proposed for MPC clustering in MIMO channel analysis, without user-specified parameters. Remarkably, the popular multipath component distance (MCD) is proved to be a special case of the proposed distance metric framework. Furthermore, two machine learning algorithms, namely, weak-supervised Mahalanobis metric for clustering and supervised large margin nearest neighbor, are introduced to learn the distance metric. To evaluate the effectiveness, a modified channel model is proposed based on the 3GPP spatial channel model to generate clustered MPCs with delay and angular information, since the original 3GPP spatial channel model (SCM) is incapable to evaluate clustering quality. Experiment results show that the proposed distance metric can significantly improve the clustering quality of existing clustering algorithms, while the learning phase requires considerably limited efforts of labeling MPCs.

cs.IT↗

arXiv2021

Multimodal Reward Shaping for Efficient Exploration in Reinforcement Learning

Maintaining the long-term exploration capability of the agent remains one of the critical challenges in deep reinforcement learning. A representative solution is to leverage reward shaping to provide intrinsic rewards for the agent to encourage exploration. However, most existing methods suffer from vanishing intrinsic rewards, which cannot provide sustainable exploration incentives. Moreover, they rely heavily on complex models and additional memory to record learning procedures, resulting in high computational complexity and low robustness. To tackle this problem, entropy-based methods are proposed to evaluate the global exploration performance, encouraging the agent to visit the state space more equitably. However, the sample complexity of estimating the state visitation entropy is prohibitive when handling environments with high-dimensional observations. In this paper, we introduce a novel metric entitled Jain's fairness index (JFI) to replace the entropy regularizer, which solves the exploration problem from a brand new perspective. In sharp contrast to the entropy regularizer, JFI is more computable and robust and can be easily applied generalized into arbitrary tasks. Furthermore, we leverage a variational auto-encoder (VAE) model to capture the life-long novelty of states, which is combined with the global JFI score to form multimodal intrinsic rewards. Finally, extensive simulation results demonstrate that our multimodal reward shaping (MMRS) method can achieve higher performance than other benchmark schemes.

cs.LG↗

arXiv2021

Analysis note: jet reconstruction, energy spectra, and substructure analyses with archived ALEPH data

The first measurements of anti-$k_{T}$ jet energy spectrum and substructure in hadronic $Z$ decays are presented. The archived $e^+e^-$ annihilation data at a center-of-mass energy of 91 GeV were collected with the ALEPH detector at LEP in 1994. The jet substructure was analyzed as a function of jet energy. The results are compared with the perturbative QCD calculations and predictions from the PYTHIA v6.1, SHERPA, and HERWIG v7.1.5 event generators. In this note, jet reconstruction procedure, jet energy calibration, and the performance with archived ALEPH data and Monte Carlo simulations are also documented.

hep-ex↗

arXiv2021

Packet Routing with Graph Attention Multi-agent Reinforcement Learning

Packet routing is a fundamental problem in communication networks that decides how the packets are directed from their source nodes to their destination nodes through some intermediate nodes. With the increasing complexity of network topology and highly dynamic traffic demand, conventional model-based and rule-based routing schemes show significant limitations, due to the simplified and unrealistic model assumptions, and lack of flexibility and adaption. Adding intelligence to the network control is becoming a trend and the key to achieving high-efficiency network operation. In this paper, we develop a model-free and data-driven routing strategy by leveraging reinforcement learning (RL), where routers interact with the network and learn from the experience to make some good routing configurations for the future. Considering the graph nature of the network topology, we design a multi-agent RL framework in combination with Graph Neural Network (GNN), tailored to the routing problem. Three deployment paradigms, centralized, federated, and cooperated learning, are explored respectively. Simulation results demonstrate that our algorithm outperforms some existing benchmark algorithms in terms of packet transmission delay and affordable load.

cs.AI↗

arXiv2021

FGLP: A Federated Fine-Grained Location Prediction System for Mobile Users

Fine-grained location prediction on smart phones can be used to improve app/system performance. Application scenarios include video quality adaptation as a function of the 5G network quality at predicted user locations, and augmented reality apps that speed up content rendering based on predicted user locations. Such use cases require prediction error in the same range as the GPS error, and no existing works on location prediction can achieve this level of accuracy. We present a system for fine-grained location prediction (FGLP) of mobile users, based on GPS traces collected on the phones. FGLP has two components: a federated learning framework and a prediction model. The framework runs on the phones of the users and also on a server that coordinates learning from all users in the system. FGLP represents the user location data as relative points in an abstract 2D space, which enables learning across different physical spaces. The model merges Bidirectional Long Short-Term Memory (BiLSTM) and Convolutional Neural Networks (CNN), where BiLSTM learns the speed and direction of the mobile users, and CNN learns information such as user movement preferences. FGLP uses federated learning to protect user privacy and reduce bandwidth consumption. Our experimental results, using a dataset with over 600,000 users, demonstrate that FGLP outperforms baseline models in terms of prediction accuracy. We also demonstrate that FGLP works well in conjunction with transfer learning, which enables model reusability. Finally, benchmark results on several types of Android phones demonstrate FGLP's feasibility in real life.

cs.LG↗

arXiv2021

Electromagnetic fields from the extended Kharzeev-McLerran-Warringa model in relativistic heavy-ion collisions

Based on the Kharzeev-McLerran-Warringa (KMW) model that estimates strong electromagnetic (EM) fields generated in relativistic heavy-ion collisions, we generalize the formulas of EM fields in the vacuum by incorporating the longitudinal position dependence, the generalized charge distributions and retardation correction. We further generalize the formulas of EM fields in the pure quark-gluon plasma (QGP) medium by incorporating a constant Ohm electric conductivity and also during the realistic early-time stages QGP evolution by using a time-dependent electric conductivity. Using the extended KMW model, we observe a slower time evolution and a more reasonable impact parameter $b$ dependence of the magnetic field strength than those from the original KMW model in the vacuum. The inclusion of medium effects by using the lattice data helps to further prolong the time evolution of magnetic field, such that the magnetic field strength during the realistic QGP evolution at thermal freeze-out time can meet the $1σ$ bound constrained from experimentally measured difference in global polarizations of $Λ$ and $\barΛ$ hyperons in Au+Au collisions at top RHIC energy. These generalized formulations in the extended KMW model will be potentially useful for many EM fields relevant studies in heavy-ion collisions, especially at lower colliding energies and for various species of colliding nuclei.

nucl-th↗

arXiv2021

Channel Measurement and Path-Loss Characterization for Low-Terahertz Indoor Scenarios

TeraHertz (THz) communications are envisioned as a promising technology, owing to its unprecedented multi-GHz bandwidth. In this paper, wideband channel measurement campaigns at 140 GHz and 220 GHz are conducted in indoor scenarios including a meeting room and an office room. Directional antennas are utilized and rotated for resolving the multi-path components (MPCs) in the angular domain. Comparable path loss values are achieved in the 140 and 220 GHz bands. To investigate the large-scale fading characteristics for indoor THz communications, single-band close-in path loss models are developed. To further analyze the dependency on the frequency, two multi-band path loss models are analyzed, i.e., alpha-beta-gamma (ABG) and multi-frequency CI model with a frequency-weighted path loss exponent (CIF), between which the ABG model demonstrates a better fit with the measured data. Moreover, a coherent beam combination that constructively sums the signal amplitudes from various arrival directions can significantly reduce the path loss, in contrast with a non-coherent beam combination.

cs.IT↗

arXiv2021

Channel Measurement and Ray-Tracing-Statistical Hybrid Modeling for Low-Terahertz Indoor Communications

TeraHertz (THz) communications are envisioned as a promising technology, owing to its unprecedented multi-GHz bandwidth. One fundamental challenge when moving to new spectrum is to understand the science of radio propagation and develop an accurate channel model. In this paper, a wideband channel measurement campaign between 130 GHz and 143 GHz is investigated in a typical meeting room. Directional antennas are utilized and rotated for resolving the multi-path components (MPCs) in the angular domain. With careful system calibration that eliminates system errors and antenna effects, a realistic power delay profile is developed. Furthermore, a combined MPC clustering and matching procedure with ray-tracing techniques is proposed to investigate the cluster behavior and wave propagation of THz signals. In light of the measurement results, physical parameters and insights in the THz indoor channel are comprehensively analyzed, including the line-of-sight path loss, power distributions, temporal and spatial features, and correlations among THz multi-path characteristics. Finally, a hybrid channel model that combines ray-tracing and statistical methods is developed for THz indoor communications. Numerical results demonstrate that the proposed hybrid channel model shows good agreement with the measurement and outperforms the conventional statistical and geometric-based stochastic channel model in terms of the temporal-spatial characteristics.

cs.IT↗

arXiv2021

A Primal-Dual Approach to Constrained Markov Decision Processes

In many operations management problems, we need to make decisions sequentially to minimize the cost while satisfying certain constraints. One modeling approach to study such problems is constrained Markov decision process (CMDP). When solving the CMDP to derive good operational policies, there are two key challenges: one is the prohibitively large state space and action space; the other is the hard-to-compute transition kernel. In this work, we develop a sampling-based primal-dual algorithm to solve CMDPs. Our approach alternatively applies regularized policy iteration to improve the policy and subgradient ascent to maintain the constraints. Under mild regularity conditions, we show that the algorithm converges at rate $ O(\log(T)/\sqrt{T})$, where T is the number of iterations. When the CMDP has a weakly coupled structure, our approach can substantially reduce the dimension of the problem through an embedded decomposition. We apply the algorithm to two important applications with weakly coupled structures: multi-product inventory management and multi-class queue scheduling, and show that it generates controls that outperform state-of-art heuristics.

math.OC↗

arXiv2021

Utilizing Dependence among Variables in Evolutionary Algorithms for Mixed-Integer Programming: A Case Study on Multi-Objective Constrained Portfolio Optimization

Several real-world applications could be modeled as Mixed-Integer Non-Linear Programming (MINLP) problems, and some prominent examples include portfolio optimization, remote sensing technology, and so on. Most of the models for these applications are non-convex and always involve some conflicting objectives. The mathematical and heuristic methods have their advantages in solving this category of problems. In this work, we turn to Multi-Objective Evolutionary Algorithms (MOEAs) for finding elegant solutions for such problems. In this framework, we investigate a multi-objective constrained portfolio optimization problem, which can be cast as a classical financial problem and can also be naturally modeled as an MINLP problem. Consequently, we point out one challenge, faced by a direct coding scheme for MOEAs, to this problem. It is that the dependence among variables, like the selection and weights for one same asset, will likely make the search difficult. We thus, propose a Compressed Coding Scheme (CCS), compressing the two dependent variables into one variable to utilize the dependence and thereby meeting this challenge. Subsequently, we carry out a detailed empirical study on two sets of instances. The first part consists of 5 instances from OR-Library, which is solvable for the general mathematical optimizer, like CPLEX, while the remaining 15 instances from NGINX are addressed only by MOEAs. The two benchmarks, involving the number of assets from 31 to 2235, consistently indicate that CCS is not only efficient but also robust for dealing with the constrained multi-objective portfolio optimization.

cs.CE↗

arXiv2021

Variable Division and Optimization for Constrained Multiobjective Portfolio Problems

Variable division and optimization (D\&O) is a frequently utilized algorithm design paradigm in Evolutionary Algorithms (EAs). A D\&O EA divides a variable into partial variables and then optimize them respectively. A complicated problem is thus divided into simple subtasks. For example, a variable of portfolio problem can be divided into two partial variables, i.e. the selection of assets and the allocation of capital. Thereby, we optimize these two partial variables respectively. There is no formal discussion about how are the partial variables iteratively optimized and why can it work for both single- and multi-objective problems in D\&O. In this paper, this gap is filled. According to the discussion, an elitist selection method for partial variables in multiobjective problems is developed. Then this method is incorporated into the Decomposition-Based Multiobjective Evolutionary Algorithm (D\&O-MOEA/D). With the help of a mathematical programming optimizer, it is achieved on the constrained multiobjective portfolio problems. In the empirical study, D\&O-MOEA/D is implemented for 20 instances and recent Chinese stock markets. The results show the superiority and versatility of D\&O-MOEA/D on large-scale instances while the performance of it on small-scale problems is also not bad. The former targets convergence towards the Pareto front and the latter helps promote diversity among the non-dominated solutions during the search process.

cs.NE↗

arXiv2020

TFGAN: Time and Frequency Domain Based Generative Adversarial Network for High-fidelity Speech Synthesis

Recently, GAN based speech synthesis methods, such as MelGAN, have become very popular. Compared to conventional autoregressive based methods, parallel structures based generators make waveform generation process fast and stable. However, the quality of generated speech by autoregressive based neural vocoders, such as WaveRNN, is still higher than GAN. To address this issue, we propose a novel vocoder model: TFGAN, which is adversarially learned both in time and frequency domain. On one hand, we propose to discriminate ground-truth waveform from synthetic one in frequency domain for offering more consistency guarantees instead of only in time domain. On the other hand, in contrast to the conventionally frequency-domain STFT loss approach or feature map loss by discriminator to learn waveform, we propose a set of time-domain loss that encourage the generator to capture the waveform directly. TFGAN has nearly same synthesis speed as MelGAN, but the fidelity is significantly improved by our novel learning method. In our experiments, TFGAN shows the ability to achieve comparable mean opinion score (MOS) than autoregressive vocoder under speech synthesis context.

eess.AS↗

arXiv2020

When Shall I Be Empathetic? The Utility of Empathetic Parameter Estimation in Multi-Agent Interactions

Human-robot interactions (HRI) can be modeled as dynamic or differential games with incomplete information, where each agent holds private reward parameters. Due to the open challenge in finding perfect Bayesian equilibria of such games, existing studies often consider approximated solutions composed of parameter estimation and motion planning steps, in order to decouple the belief and physical dynamics. In parameter estimation, current approaches often assume that the reward parameters of the robot are known by the humans. We argue that by falsely conditioning on this assumption, the robot performs non-empathetic estimation of the humans' parameters, leading to undesirable values even in the simplest interactions. We test this argument by studying a two-vehicle uncontrolled intersection case with short reaction time. Results show that when both agents are unknowingly aggressive (or non-aggressive), empathy leads to more effective parameter estimation and higher reward values, suggesting that empathy is necessary when the true parameters of agents mismatch with their common belief. The proposed estimation and planning algorithms are therefore more robust than the existing approaches, by fully acknowledging the nature of information asymmetry in HRI. Lastly, we introduce value approximation techniques for real-time execution of the proposed algorithms.

cs.RO↗

arXiv2020

Chiral Triclinic Metamaterial Crystals Supporting Isotropic Acoustical Activity and Isotropic Chiral Phonons

Recent work predicted the existence of isotropic chiral phonon dispersion relations of the lowest bands connected to isotropic acoustical activity in cubic crystalline approximants of 3D chiral icosahedral metamaterial quasicrystals. While these architectures are fairly broadband and presumably robust against fabrication tolerances due to orientation averaging, they are extremely complex, very hard to manufacture experimentally, and they show effects which are about an order of magnitude smaller compared to those of ordinary highly anisotropic chiral cubic metamaterial crystals. Here, we propose and analyze a chiral triclinic metamaterial crystal exhibiting broadband isotropic acoustical activity. These 3D truss lattices are much less complex and exhibit substantially larger effects than the 3D quasicrystals at the price of being somewhat more susceptible to fabrication tolerances. This susceptibility originates from the fact that we have tailored the lowest two transverse phonon bands to exhibit an "accidental" degeneracy in momentum space.

cond-mat.mtrl-sci↗

arXiv2020

Imaging spinon density modulations in a 2D quantum spin liquid

Two-dimensional triangular-lattice antiferromagnets are predicted under some conditions to exhibit a quantum spin liquid ground state whose low-energy behavior is described by a spinon Fermi surface. Directly imaging the resulting spinons, however, is difficult due to their fractional, chargeless nature. Here we use scanning tunneling spectroscopy to image spinon density modulations arising from a spinon Fermi surface instability in single-layer 1T-TaSe$_2$, a two-dimensional Mott insulator. We first demonstrate the existence of localized spins arranged on a triangular lattice in single-layer 1T-TaSe$_2$ by contacting it to a metallic 1H-TaSe$_2$ layer and measuring the Kondo effect. Subsequent spectroscopic imaging of isolated, single-layer 1T-TaSe$_2$ reveals long-wavelength modulations at Hubbard band energies that reflect spinon density modulations. This allows direct experimental measurement of the spinon Fermi wavevector, in good agreement with theoretical predictions for a 2D quantum spin liquid. These results establish single-layer 1T-TaSe$_2$ as a new platform for studying novel two-dimensional quantum-spin-liquid phenomena.

cond-mat.str-el↗

arXiv2020

Reverse-engineering Bar Charts Using Neural Networks

Reverse-engineering bar charts extracts textual and numeric information from the visual representations of bar charts to support application scenarios that require the underlying information. In this paper, we propose a neural network-based method for reverse-engineering bar charts. We adopt a neural network-based object detection model to simultaneously localize and classify textual information. This approach improves the efficiency of textual information extraction. We design an encoder-decoder framework that integrates convolutional and recurrent neural networks to extract numeric information. We further introduce an attention mechanism into the framework to achieve high accuracy and robustness. Synthetic and real-world datasets are used to evaluate the effectiveness of the method. To the best of our knowledge, this work takes the lead in constructing a complete neural network-based method of reverse-engineering bar charts.

cs.CV↗