arXiv subjects

Jun Zhang

Publications and source records attributed to Jun Zhang.

At least 793 records · Page 44Linked to original sources

arXiv2021

RBUE: A ReLU-Based Uncertainty Estimation Method of Deep Neural Networks

Deep neural networks (DNNs) have successfully learned useful data representations in various tasks. However, assessing the reliability of these representations remains a challenge. Deep Ensemble is widely considered the state-of-the-art method which can estimate the uncertainty with higher quality, but it is very expensive to train and test. MC-Dropout is another popular method, which is less expensive but lacks the diversity of predictions. To estimate the uncertainty with higher quality in less time, we introduce a ReLU-Based Uncertainty Estimation (RBUE) method. Instead of randomly dropping some neurons of the network as in MC-Dropout or using the randomness of the initial weights of networks as in Deep Ensemble, RBUE adds randomness to the activation function module, making the outputs diverse. Under the method, we propose two strategies, MC-DropReLU and MC-RReLU, to estimate uncertainty. We analyze and compare the output diversity of MC-Dropout and our method from the variance perspective and obtain the relationship between the hyperparameters and predictive diversity in the two methods. Moreover, our method is simple to implement and does not need to modify the existing model. We experimentally validate the RBUE on three widely used datasets, CIFAR10, CIFAR100, and TinyImageNet. The experiments demonstrate that our method has competitive performance but is more favorable in training time and memory requirements.

cs.LG↗

arXiv2021

Practical quantum access network over a 10 Gbit/s Ethernet passive optical network

Quantum key distribution (QKD) provides an information-theoretically secure method to share keys between legitimate users. To achieve large-scale deployment of QKD, it should be easily scalable and cost-effective. The infrastructure construction of quantum access network (QAN) expands network capacity and the integration between QKD and classical optical communications reduces the cost of channel. Here, we present a practical downstream QAN over a 10 Gbit/s Ethernet passive optical network (10G-EPON), which can support up to 64 users. In the full coexistence scheme using the single feeder fiber structure, the co-propagation of QAN and 10G-EPON signals with 9 dB attenuation is achieved over 21 km fiber, and the secure key rate for each of 16 users reaches 1.5 kbps. In the partial coexistence scheme using the dual feeder fiber structure, the combination of QAN and full-power 10G-EPON signals is achieved over 11 km with a network capacity of 64-user. The practical QAN over the 10G-EPON in our work implements an important step towards the achievement of large-scale QKD infrastructure.

quant-ph↗

arXiv2021

Detected the steerability bounds of the generalized Werner states via BackPropagation neural network

We use error BackPropagation (BP) neural network to determine whether an arbitrary two-qubit quantum state is steerable and optimize the steerability bounds of the generalized Werner state. The results show that no matter how we choose the features for the quantum states, we can use the BP neural network to construct several models to realize high-performance quantum steering classifiers compared with the support vector machine (SVM). In addition, we predict the steerability bounds of the generalized Werner states by using the classifiers which are newly constructed by the BP neural network, that is, the predicted steerability bounds are closer to the theoretical bounds. In particular, high-performance classifiers with partial information of the quantum states which we only need to measure in three fixed measurement directions are obtained.

quant-ph↗

arXiv2021

Sk-Unet Model with Fourier Domain for Mitosis Detection

Mitotic count is the most important morphological feature of breast cancer grading. Many deep learning-based methods have been proposed but suffer from domain shift. In this work, we construct a Fourier-based segmentation model for mitosis detection to address the problem. Swapping the low-frequency spectrum of source and target images is shown effective to alleviate the discrepancy between different scanners. Our Fourier-based segmentation method can achieve F1 with 0.7456 on the preliminary test set.

eess.IV↗

arXiv2021

Gradient zigzag metamaterial beams as broadband vibration isolators for beam-like structures

Phononic crystals (PCs) and metamaterials are artificially structured materials with the unprecedented property of the existence of complete stop-bands. However, the generally narrow width of the stop-bands in such materials severely limits their applicability. Here, a particular kind of gradient zigzag metamaterial beams exclusively with flat bands within a broad frequency range is proposed based upon our previously proposed uniform counterparts. The calculated band structures by the Finite Element Method (FEM) and the Spectral Element Method (SEM) show the distinctly different characteristics of the gradient zigzag metamaterial beams from the conventional PCs and metamaterials, where the band structures are totally filled with flat bands within a broad frequency range without the normal pass bands. These flat bands are caused by the local resonance of certain parts in the gradient zigzag metamaterial beams, and the elastic waves with the frequencies at these flat bands are trapped in the resonant parts and cannot propagate. As a result, our proposed gradient zigzag metamaterial beams can be used as broadband vibration isolators for beam-like structures, which is verified by both numerical simulations and experiments conducted on the 3D-printed samples. This design strategy opens a new avenue for designing and constructing novel broadband vibration isolators.

physics.app-ph↗

arXiv2021

Destabilization of black holes and stars by generalized Proca fields

We demonstrate that black holes and stars in general relativity can be destabilized by perturbations of non-minimally coupled vector fields. Focusing on static and spherically symmetric backgrounds, our analysis shows that black holes with sufficiently small mass and stars with sufficiently high densities are subject to ghost or gradient-type instabilities. This holds for a large class of vector-tensor theories whenever non-minimal couplings contribute to linearized dynamics about a state with vanishing vector field and applies to generalized Proca models that have sparked attention for their potential role in cosmology and astrophysics. The stability criteria translate into bounds of relevance for low-scale theories of dark energy and for ultra-light dark matter scenarios.

gr-qc↗

arXiv2021

Device-independent randomness expansion against quantum side information

The ability to produce random numbers that are unknown to any outside party is crucial for many applications. Device-independent randomness generation does not require trusted devices and therefore provides strong guarantees of the security of the output, but it comes at the price of requiring the violation of a Bell inequality for implementation. A further challenge is to make the bounds in the security proofs tight enough to allow randomness expansion with contemporary technology. Although randomness has been generated in recent experiments, the amount of randomness consumed in doing so has been too high to certify expansion based on existing theory. Here we present an experiment that demonstrates device-independent randomness expansion. By developing a Bell test setup with a single-photon detection efficiency of around $84\%$ and by using a spot-checking protocol, we achieve a net gain of $2.57\times10^8$ certified bits with a soundness error $3.09\times10^{-12}$. The experiment ran for $19.2$ h, which corresponds to an average rate of randomness generation of $13,527$ bits per second. By developing the entropy accumulation theorem, we establish security against quantum adversaries. We anticipate that this work will lead to further improvements that push device-independence towards commercial viability.

quant-ph↗

arXiv2021

Learn to Communicate with Neural Calibration: Scalability and Generalization

The conventional design of wireless communication systems typically relies on established mathematical models that capture the characteristics of different communication modules. Unfortunately, such design cannot be easily and directly applied to future wireless networks, which will be characterized by large-scale ultra-dense networks whose design complexity scales exponentially with the network size. Furthermore, such networks will vary dynamically in a significant way, which makes it intractable to develop comprehensive analytical models. Recently, deep learning-based approaches have emerged as potential alternatives for designing complex and dynamic wireless systems. However, existing learning-based methods have limited capabilities to scale with the problem size and to generalize with varying network settings. In this paper, we propose a scalable and generalizable neural calibration framework for future wireless system design, where a neural network is adopted to calibrate the input of conventional model-based algorithms. Specifically, the backbone of a traditional time-efficient algorithm is integrated with deep neural networks to achieve a high computational efficiency, while enjoying enhanced performance. The permutation equivariance property, carried out by the topological structure of wireless systems, is furthermore utilized to develop a generalizable neural network architecture. The proposed neural calibration framework is applied to solve challenging resource management problems in massive multiple-input multiple-output (MIMO) systems. Simulation results will show that the proposed neural calibration approach enjoys significantly improved scalability and generalization compared with the existing learning-based methods.

eess.SP↗

arXiv2021

An Open Problem on the Bentness of Mesnager's Functions

Let $n=2m$. In the present paper, we study the binomial Boolean functions of the form $$f_{a,b}(x) = \mathrm{Tr}_1^{n}(a x^{2^m-1 }) +\mathrm{Tr}_1^{2}(bx^{\frac{2^n-1}{3} }), $$ where $m$ is an even positive integer, $a\in \mathbb{F}_{2^n}^*$ and $b\in \mathbb{F}_4^*$. We show that $ f_{a,b}$ is a bent function if the Kloosterman sum $$K_{m}\left(a^{2^m+1}\right)=1+ \sum_{x\in \mathbb{F}_{2^m}^*} (-1)^{\mathrm{Tr}_1^{m}(a^{2^m+1} x+ \frac{1}{x})}$$ equals $4$, thus settling an open problem of Mesnager. The proof employs tools including computing Walsh coefficients of Boolean functions via multiplicative characters, divisibility properties of Gauss sums, and graph theory.

cs.IT↗

arXiv2021

An accurate moving wall boundary algorithm for Direct Simulation of Monte Carlo in unsteady rarefied flow

An accurate algorithm is proposed to improve the prediction of a particle in collision with a moving wall within the direct simulation Monte Carlo (DSMC) framework for the simulation of unsteady rarefied flows. This algorithm is able to predict the particle-wall collision in a coupled manner by removing the assumption employed by the approximate algorithm, in that the wall is frozen during the collision. The trajectory equation of the particle is theoretically constructed in a moving object coordinate system. It can accurately describe the geometries of the collision between a particle and an arbitrarily shaped object of which the motion incorporates both translation and rotation, thus allowing to deal with complex problems. In contrast, the approximate algorithm ignores the effect of the moving wall on the particle movement during the collision, and therefore induces error that is an increasing function of the wall velocity. Four rarefied flow problems are applied to validate the accurate algorithm. It is shown that the algorithm can produce results perfectly consistent with the Maxwellian theoretical solutions and ensure particle conservation to avoid gas leakage. It is also shown in a three-dimensional case of a re-entry module that the steady simulation fails to reproduce the hysteresis effect while the unsteady simulation using the accurate algorithm can do that, indicating that the unsteady simulation with an appropriate algorithm as proposed in the present work is essentially required in such applications.

physics.comp-ph↗

arXiv2021

First Constraints on Nuclear Coupling of Axionlike Particles from the Binary Neutron Star Gravitational Wave Event GW170817

Light axion fields, if they exist, can be sourced by neutron stars due to their coupling to nuclear matter, and play a role in binary neutron star mergers. We report on a search for such axions by analysing the gravitational waves from the binary neutron star inspiral GW170817. We find no evidence of axions in the sampled parameter space. The null result allows us to impose constraints on axions with masses below $10^{-11} {\rm eV}$ by excluding the ones with decay constants ranging from $1.6\times10^{16} {\rm GeV}$ to $10^{18} {\rm GeV}$ at $3σ$ confidence level. Our analysis provides the first constraints on axions from neutron star inspirals, and rules out a large region in parameter space that has not been probed by the existing experiments.

hep-ph↗

arXiv2021

Deriving the PEE proposal from the Locking bit thread configuration

In the holographic framework, we argue that the partial entanglement entropy (PEE) can be explicitly interpreted as the component flow flux in a locking bit thread configuration. By applying the locking theorem of bit threads, and constructing a concrete locking scheme, we obtain a set of uniquely determined component flow fluxes from this viewpoint, and successfully derive the PEE proposal and its generalized version in the multipartite cases. Moreover, from this perspective of bit threads, we also present a coherent explanation for the coincidence between the BPE (balanced partial entanglement)/EWCS (entanglement wedge cross section) duality proposed recently and the EoP (entanglement of purification)/EWCS duality. We also discuss the issues implied by this coincident between the idea of the PEE and the picture of locking thread configuration.

hep-th↗

arXiv2021

Enhanced heat transport in thermal convection with suspensions of rod-like expandable particles

Thermal convection of fluid is a more efficient way than diffusion to carry heat from hot sources to cold places. Here, we experimentally study the Rayleigh-Bénard convection of aqueous glycerol solution in a cubic cell with suspensions of rod-like particles made of polydimethylsiloxane (PDMS). The particles are inertial due to their large thermal expansion coefficient and finite sizes. The thermal expansion coefficient of the particles is three times larger than that of the background fluid. This contrast makes the suspended particles lighter than the local fluid in hot regions and heavier in cold regions. The heat transport is enhanced at relatively large Rayleigh number ($Ra$) but reduced at small $Ra$. We demonstrate that the increase of Nusselt number arises from the particle-boundary layer interactions: the particles act as ``active'' mixers of the flow and temperature fields across the boundary layers.

physics.flu-dyn↗

arXiv2021

Communication-Computation Efficient Device-Edge Co-Inference via AutoML

Device-edge co-inference, which partitions a deep neural network between a resource-constrained mobile device and an edge server, recently emerges as a promising paradigm to support intelligent mobile applications. To accelerate the inference process, on-device model sparsification and intermediate feature compression are regarded as two prominent techniques. However, as the on-device model sparsity level and intermediate feature compression ratio have direct impacts on computation workload and communication overhead respectively, and both of them affect the inference accuracy, finding the optimal values of these hyper-parameters brings a major challenge due to the large search space. In this paper, we endeavor to develop an efficient algorithm to determine these hyper-parameters. By selecting a suitable model split point and a pair of encoder/decoder for the intermediate feature vector, this problem is casted as a sequential decision problem, for which, a novel automated machine learning (AutoML) framework is proposed based on deep reinforcement learning (DRL). Experiment results on an image classification task demonstrate the effectiveness of the proposed framework in achieving a better communication-computation trade-off and significant inference speedup against various baseline schemes.

cs.LG↗

arXiv2021

Image-free single-pixel segmentation

The existing segmentation techniques require high-fidelity images as input to perform semantic segmentation. Since the segmentation results contain most of edge information that is much less than the acquired images, the throughput gap leads to both hardware and software waste. In this letter, we report an image-free single-pixel segmentation technique. The technique combines structured illumination and single-pixel detection together, to efficiently samples and multiplexes scene's segmentation information into compressed one-dimensional measurements. The illumination patterns are optimized together with the subsequent reconstruction neural network, which directly infers segmentation maps from the single-pixel measurements. The end-to-end encoding-and-decoding learning framework enables optimized illumination with corresponding network, which provides both high acquisition and segmentation efficiency. Both simulation and experimental results validate that accurate segmentation can be achieved using two-order-of-magnitude less input data. When the sampling ratio is 1%, the Dice coefficient reaches above 80% and the pixel accuracy reaches above 96%. We envision that this image-free segmentation technique can be widely applied in various resource-limited platforms such as UAV and unmanned vehicle that require real-time sensing.

eess.IV↗

arXiv2021

Compressive Sensing Based Joint Activity and Data Detection for Grant-Free Massive IoT Access

Massive machine-type communications (mMTC) are poised to provide ubiquitous connectivity for billions of Internet-of-Things (IoT) devices. However, the required low-latency massive access necessitates a paradigm shift in the design of random access schemes, which invokes a need of efficient joint activity and data detection (JADD) algorithms. By exploiting the feature of sporadic traffic in massive access, a beacon-aided slotted grant-free massive access solution is proposed. Specifically, we spread the uplink access signals in multiple subcarriers with pre-equalization processing and formulate the JADD as a multiple measurement vector (MMV) compressive sensing problem. Moreover, to leverage the structured sparsity of uplink massive access signals among multiple time slots, we develop two computationally efficient detection algorithms, which are termed as orthogonal approximate message passing (OAMP)-MMV algorithm with simplified structure learning (SSL) and accurate structure learning (ASL). To achieve accurate detection, the expectation maximization algorithm is exploited for learning the sparsity ratio and the noise variance. To further improve the detection performance, channel coding is applied and successive interference cancellation (SIC)-based OAMP-MMV-SSL and OAMP-MMV-ASL algorithms are developed, where the likelihood ratio obtained in the soft-decision can be exploited for refining the activity identification. Finally, the state evolution of the proposed OAMP-MMV-SSL and OAMP-MMV-ASL algorithms is derived to predict the performance theoretically. Simulation results verify that the proposed solutions outperform various state-of-the-art baseline schemes, enabling low-latency random access and high-reliable massive IoT connectivity with overloading.

cs.IT↗

arXiv2021

Distributed Expectation Propagation Detection for Cell-Free Massive MIMO

In cell-free massive MIMO networks, an efficient distributed detection algorithm is of significant importance. In this paper, we propose a distributed expectation propagation (EP) detector for cell-free massive MIMO. The detector is composed of two modules, a nonlinear module at the central processing unit (CPU) and a linear module at the access point (AP). The turbo principle in iterative decoding is utilized to compute and pass the extrinsic information between modules. An analytical framework is then provided to characterize the asymptotic performance of the proposed EP detector with a large number of antennas. Simulation results will show that the proposed method outperforms the distributed detectors in terms of bit-error-rate.

cs.IT↗

arXiv2021

Semantic Reinforced Attention Learning for Visual Place Recognition

Large-scale visual place recognition (VPR) is inherently challenging because not all visual cues in the image are beneficial to the task. In order to highlight the task-relevant visual cues in the feature embedding, the existing attention mechanisms are either based on artificial rules or trained in a thorough data-driven manner. To fill the gap between the two types, we propose a novel Semantic Reinforced Attention Learning Network (SRALNet), in which the inferred attention can benefit from both semantic priors and data-driven fine-tuning. The contribution lies in two-folds. (1) To suppress misleading local features, an interpretable local weighting scheme is proposed based on hierarchical feature distribution. (2) By exploiting the interpretability of the local weighting scheme, a semantic constrained initialization is proposed so that the local attention can be reinforced by semantic priors. Experiments demonstrate that our method outperforms state-of-the-art techniques on city-scale VPR benchmark datasets.

cs.CV↗