arXiv subjects

Yu Zhou

Publications and source records attributed to Yu Zhou.

At least 325 records · Page 18Linked to original sources

arXiv2022

MARF: Multiscale Adaptive-switch Random Forest for Leg Detection with 2D Laser Scanners

For the 2D laser-based tasks, e.g., people detection and people tracking, leg detection is usually the first step. Thus, it carries great weight in determining the performance of people detection and people tracking. However, many leg detectors ignore the inevitable noise and the multiscale characteristics of the laser scan, which makes them sensitive to the unreliable features of point cloud and further degrades the performance of the leg detector. In this paper, we propose a multiscale adaptive-switch Random Forest (MARF) to overcome these two challenges. Firstly, the adaptive-switch decision tree is designed to use noisesensitive features to conduct weighted classification and noiseinvariant features to conduct binary classification, which makes our detector perform more robust to noise. Secondly, considering the multiscale property that the sparsity of the 2D point cloud is proportional to the length of laser beams, we design a multiscale random forest structure to detect legs at different distances. Moreover, the proposed approach allows us to discover a sparser human leg from point clouds than others. Consequently, our method shows an improved performance compared to other state-of-the-art leg detectors on the challenging Moving Legs dataset and retains the whole pipeline at a speed of 60+ FPS on lowcomputational laptops. Moreover, we further apply the proposed MARF to the people detection and tracking system, achieving a considerable gain in all metrics.

cs.RO↗

arXiv2022

Scaling Bockchain with Adaptivity

This paper presents Balloon, a scalable blockchain consensus protocol which could dynamically adapt its performance to the overall computation power change. Balloon is based on a parallel chain architecture combined with a greedy heaviest sub-chain selection strategy. It adopts an inovative block sampling approach to assess the change of block generation rate in the network. By introducing view change mechanism, Balllon is able to dynamically adjust the number of parallel sub-chains. Balloon redefines the concept of block subtree weight with view change in consideration, so that a total order of blocks could be obtained safely. To deal with rapidly increasing block generation rate in the blockchain network, participants of previous Nakamoto-style protocols are required to continuously increase their mining difficulty so as to maintain an expected security gurantee. Balloon, however, could accomadate a fixed difficulty setup and assign superfluous block processing capability to new sub-chains, which makes it more open and also economical.

cs.DC↗

arXiv2022

Transformation-Invariant Learning of Optimal Individualized Decision Rules with Time-to-Event Outcomes

In many important applications of precision medicine, the outcome of interest is time to an event (e.g., death, relapse of disease) and the primary goal is to identify the optimal individualized decision rule (IDR) to prolong survival time. Existing work in this area have been mostly focused on estimating the optimal IDR to maximize the We propose a new robust framework for estimating an optimal static or dynamic IDR with time-to-event outcomes based on an easy-to-interpret quantile criterion. The new method does not need to specify an outcome regression model and is robust for heavy-tailed distribution. The estimation problem corresponds to a nonregular M-estimation problem with both finite and infinite-dimensional nuisance parameters. Employing advanced empirical process techniques, we establish the statistical theory of the estimated parameter indexing the optimal IDR. Furthermore, we prove a novel result that the proposed approach can consistently estimate the optimal value function under mild conditions even when the optimal IDR is non-unique, which happens in the challenging setting of exceptional laws. We also propose a smoothed resampling procedure for inference. The proposed methods are implemented in the R-package QTOCen. We demonstrate the performance of the proposed new methods via extensive Monte Carlo studies and a real data application.restricted mean survival time in the population.

stat.ME↗

arXiv2022

Investigations into the characteristics and influences of nonequilibrium evolution

In order to estimate qualitatively the influence of nonequilibrium evolution in relativistic heavy ion collisions, we use the three dimensional Ising model with Metropolis algorithm to study the evolution from nonequilibrium to equilibrium on the phase boundary. The evolution of order parameter approaches its equilibrium value exponentially, the same as that given by Langevin equation. The average relaxation time is defined which is demonstrated to well represent the relaxation time in dynamical equations. It is shown that the average relaxation time at critical temperature diverges as the zth power of system size. The third and the fourth cumulants of order parameter during the nonequilibrium evolution could be either positive or negative, depending on the observation time, consistent with dynamical models at T > Tc. It is found that the nonequilibrium evolution at T > Tc lasts very short, and the influence is weaker than that at T < Tc. Those qualitative features are instructive to determine experimentally the critical point and the phase boundary of QCD.

cond-mat.stat-mech↗

arXiv2022

Towards Escaping from Language Bias and OCR Error: Semantics-Centered Text Visual Question Answering

Texts in scene images convey critical information for scene understanding and reasoning. The abilities of reading and reasoning matter for the model in the text-based visual question answering (TextVQA) process. However, current TextVQA models do not center on the text and suffer from several limitations. The model is easily dominated by language biases and optical character recognition (OCR) errors due to the absence of semantic guidance in the answer prediction process. In this paper, we propose a novel Semantics-Centered Network (SC-Net) that consists of an instance-level contrastive semantic prediction module (ICSP) and a semantics-centered transformer module (SCT). Equipped with the two modules, the semantics-centered model can resist the language biases and the accumulated errors from OCR. Extensive experiments on TextVQA and ST-VQA datasets show the effectiveness of our model. SC-Net surpasses previous works with a noticeable margin and is more reasonable for the TextVQA task.

cs.CV↗

arXiv2022

Dispersion Analysis of CIP-FEM for Helmholtz Equation

When solving the Helmholtz equation numerically, the accuracy of numerical solution deteriorates as the wave number $k$ increases, known as `pollution effect' which is directly related to the phase difference between the exact and numerical solutions, caused by the numerical dispersion. In this paper, we propose a dispersion analysis for the continuous interior penalty finite element method (CIP-FEM) and derive an explicit formula of the penalty parameter for the $p^{\rm th}$ order CIP-FEM on tensor product (Cartesian) meshes, with which the phase difference is reduced from $\mathcal{O}\big(k(kh)^{2p}\big)$ to $\mathcal{O}\big(k(kh)^{2p+2}\big)$. Extensive numerical tests show that the pollution error of the CIP-FE solution is also reduced by two orders in $kh$ with the same penalty parameter.

math.NA↗

arXiv2022

Spatial Power Spectral Analysis of the Suzaku X-ray Background

Power spectra of spatial fluctuations of X-ray emission may impose constraints on the origins of the emission independent of that from the energy spectra. We generated spatial power spectrum densities (PSD) of blank X-ray skies observed with Suzaku X-ray observatory utilizing the modified $Δ$-variance method. Using the total measured count rate as the diagnostic tool, we found that a model consisting of the sum of two components, one for the unresolved faint point sources and one for the uniform flat-field emission, can well represent the observed PSD in three different energy bands (0.2-0.5 keV, 0.5-2 keV, and 2-10 keV); only an upper limit is obtained for the latter component in 2-10 keV. X-ray counting rates corresponding to the best-fit PSD model functions and diffuse emission fractions were estimated, and we confirmed that the sum of the counting rates of two model components is consistent with those actually observed with the detector for all energy bands. The ratio of the flat-field counting rate to the total in 0.5-2 keV, however, is significantly larger than the diffuse emission fraction estimated from the model fits of energy spectra. We discussed that this discrepancy can be reconciled by systematic effects in the PSD and energy spectrum analyses. The present study demonstrates that the spatial power spectrum is powerful in constraining the origins of the X-ray emission.

astro-ph.HE↗

arXiv2022

Graph Neural Networks: Taxonomy, Advances and Trends

Graph neural networks provide a powerful toolkit for embedding real-world graphs into low-dimensional spaces according to specific tasks. Up to now, there have been several surveys on this topic. However, they usually lay emphasis on different angles so that the readers can not see a panorama of the graph neural networks. This survey aims to overcome this limitation, and provide a comprehensive review on the graph neural networks. First of all, we provide a novel taxonomy for the graph neural networks, and then refer to up to 400 relevant literatures to show the panorama of the graph neural networks. All of them are classified into the corresponding categories. In order to drive the graph neural networks into a new stage, we summarize four future research directions so as to overcome the facing challenges. It is expected that more and more scholars can understand and exploit the graph neural networks, and use them in their research community.

cs.LG↗

arXiv2022

Imagine by Reasoning: A Reasoning-Based Implicit Semantic Data Augmentation for Long-Tailed Classification

Real-world data often follows a long-tailed distribution, which makes the performance of existing classification algorithms degrade heavily. A key issue is that samples in tail categories fail to depict their intra-class diversity. Humans can imagine a sample in new poses, scenes, and view angles with their prior knowledge even if it is the first time to see this category. Inspired by this, we propose a novel reasoning-based implicit semantic data augmentation method to borrow transformation directions from other classes. Since the covariance matrix of each category represents the feature transformation directions, we can sample new directions from similar categories to generate definitely different instances. Specifically, the long-tailed distributed data is first adopted to train a backbone and a classifier. Then, a covariance matrix for each category is estimated, and a knowledge graph is constructed to store the relations of any two categories. Finally, tail samples are adaptively enhanced via propagating information from all the similar categories in the knowledge graph. Experimental results on CIFAR-100-LT, ImageNet-LT, and iNaturalist 2018 have demonstrated the effectiveness of our proposed method compared with the state-of-the-art methods.

cs.CV↗

arXiv2021

Nth-order nonlinear intensity fluctuation amplifier

Stronger light intensity fluctuations are pursued by related applications such as optical resolution, image enhancement, and beam positioning. In this paper, an Nth-order light intensity fluctuation amplifier is proposed, which was demonstrated by a four-wave mixing process with different statistical distribution coupling lights. Firstly, its amplification mechanism is revealed both theoretically and experimentally. The ratio $R$ of statistical distributions and the degree of second-order coherence ${g^{(2)}}(0)$ of beams are used to characterize the affected modulations and the increased light intensity fluctuations through the four-wave mixing process. The results show that the amplification of light intensity fluctuations is caused by not only the fluctuating light fields of incident coupling beams, but also the fluctuating nonlinear coefficient of interaction. At last, we highlight the potentiality of applying such amplifier to other N-order nonlinear optical effects.

physics.optics↗

arXiv2021

Video Text Tracking With a Spatio-Temporal Complementary Model

Text tracking is to track multiple texts in a video,and construct a trajectory for each text. Existing methodstackle this task by utilizing the tracking-by-detection frame-work, i.e., detecting the text instances in each frame andassociating the corresponding text instances in consecutiveframes. We argue that the tracking accuracy of this paradigmis severely limited in more complex scenarios, e.g., owing tomotion blur, etc., the missed detection of text instances causesthe break of the text trajectory. In addition, different textinstances with similar appearance are easily confused, leadingto the incorrect association of the text instances. To this end,a novel spatio-temporal complementary text tracking model isproposed in this paper. We leverage a Siamese ComplementaryModule to fully exploit the continuity characteristic of the textinstances in the temporal dimension, which effectively alleviatesthe missed detection of the text instances, and hence ensuresthe completeness of each text trajectory. We further integratethe semantic cues and the visual cues of the text instance intoa unified representation via a text similarity learning network,which supplies a high discriminative power in the presence oftext instances with similar appearance, and thus avoids the mis-association between them. Our method achieves state-of-the-art performance on several public benchmarks. The source codeis available at https://github.com/lsabrinax/VideoTextSCM.

cs.CV↗

arXiv2021

Anomaly Discovery in Semantic Segmentation via Distillation Comparison Networks

This paper aims to address the problem of anomaly discovery in semantic segmentation. Our key observation is that semantic classification plays a critical role in existing approaches, while the incorrectly classified pixels are easily regarded as anomalies. Such a phenomenon frequently appears and is rarely discussed, which significantly reduces the performance of anomaly discovery. To this end, we propose a novel Distillation Comparison Network (DiCNet). It comprises of a teacher branch which is a semantic segmentation network that removed the semantic classification head, and a student branch that is distilled from the teacher branch through a distribution distillation. We show that the distillation guarantees the semantic features of the two branches hold consistency in the known classes, while reflect inconsistency in the unknown class. Therefore, we leverage the semantic feature discrepancy between the two branches to discover the anomalies. DiCNet abandons the semantic classification head in the inference process, and hence significantly alleviates the issue caused by incorrect semantic classification. Extensive experimental results on StreetHazards dataset and BDD-Anomaly dataset are conducted to verify the superior performance of DiCNet. In particular, DiCNet obtains a 6.3% improvement in AUPR and a 5.2% improvement in FPR95 on StreetHazards dataset, achieves a 4.2% improvement in AUPR and a 6.8% improvement in FPR95 on BDD-Anomaly dataset. Codes are available at https://github.com/zhouhuan-hust/DiCNet.

cs.CV↗

arXiv2021

On support $τ$-tilting graphs of gentle algebras

Let $A$ be a finite-dimensional gentle algebra over an algebraically closed field. We investigate the combinatorial properties of support $τ$-tilting graph of $A$. In particular, it is proved that the support $τ$-tilting graph of $A$ is connected and has the so-called reachable-in-face property. This property was conjectured by Fomin and Zelevinsky for exchange graphs of cluster algebras which was recently confirmed by Cao and Li.

math.RT↗

arXiv2021

Learning to Select the Next Reasonable Mention for Entity Linking

Entity linking aims to establish a link between entity mentions in a document and the corresponding entities in knowledge graphs (KGs). Previous work has shown the effectiveness of global coherence for entity linking. However, most of the existing global linking methods based on sequential decisions focus on how to utilize previously linked entities to enhance the later decisions. In those methods, the order of mention is fixed, making the model unable to adjust the subsequent linking targets according to the previously linked results, which will cause the previous information to be unreasonably utilized. To address the problem, we propose a novel model, called DyMen, to dynamically adjust the subsequent linking target based on the previously linked entities via reinforcement learning, enabling the model to select a link target that can fully use previously linked information. We sample mention by sliding window to reduce the action sampling space of reinforcement learning and maintain the semantic coherence of mention. Experiments conducted on several benchmark datasets have shown the effectiveness of the proposed model.

cs.CL↗

arXiv2021

Ghost Imaging Based on Recurrent Neural Network

Benefit from the promising features of second-order correlation, ghost imaging (GI) has received extensive attentions in recent years. Simultaneously, GI is affected by the poor trade-off between sampling rate and imaging quality. The traditional image reconstruction method in GI is to accumulate the action result of each speckle and the corresponding bucket signal. We found that the image reconstruction process of GI is very similar to the Recurrent Neural Network (RNN), which is one of the deep learning algorithm. In this paper, we proposed a novel method that effectively implements GI on the RNN architecture, called GI-RNN. The state of each layer in RNN is determined by the output of the previous layer and the input of this layer, and the output of the network is the sum of all previous states. Therefore, we take the speckle of each illumination and the corresponding bucket signal as the input of each layer, and the output of the network is the sum of all previous speckle and bucket signal, which is the image of the target. The testing results show that the proposed method can achieve image reconstruction at a very low sampling rate (0.38$\%$). Moreover, we compare GI-RNN with traditional GI algorithm and compressed sensing algorithm. The results of different targets show that GI-RNN is 6.61 dB higher than compressed sensing algorithm and 12.58 dB higher than traditional GI algorithm on average. In our view, the proposed method makes an important step to applications of GI.

eess.IV↗

arXiv2021

Adversarial Robustness of Deep Code Comment Generation

Deep neural networks (DNNs) have shown remarkable performance in a variety of domains such as computer vision, speech recognition, or natural language processing. Recently they also have been applied to various software engineering tasks, typically involving processing source code. DNNs are well-known to be vulnerable to adversarial examples, i.e., fabricated inputs that could lead to various misbehaviors of the DNN model while being perceived as benign by humans. In this paper, we focus on the code comment generation task in software engineering and study the robustness issue of the DNNs when they are applied to this task. We propose ACCENT, an identifier substitution approach to craft adversarial code snippets, which are syntactically correct and semantically close to the original code snippet, but may mislead the DNNs to produce completely irrelevant code comments. In order to improve the robustness, ACCENT also incorporates a novel training method, which can be applied to existing code comment generation models. We conduct comprehensive experiments to evaluate our approach by attacking the mainstream encoder-decoder architectures on two large-scale publicly available datasets. The results show that ACCENT efficiently produces stable attacks with functionality-preserving adversarial examples, and the generated examples have better transferability compared with baselines. We also confirm, via experiments, the effectiveness in improving model robustness with our training method.

cs.SE↗

arXiv2021

Weak cotorsion, $τ$-tilting and two-term categories

Motivated by its links to $τ$-tilting theory, we introduce a generalization of cotorsion pairs in module categories. Such pairs are also linked to co-t-structures in corresponding triangulated categories, and to cotorsion pairs in certain extension closed (and hence extriangulated) subcategories, which we call two-term categories.

math.RT↗

arXiv2021

Rapid and Unconditional Parametric Reset Protocol for Tunable Superconducting Qubits

Qubit initialization is a critical task in quantum computation and communication. Extensive efforts have been made to achieve this with high speed, efficiency and scalability. However, previous approaches have either been measurement-based and required fast feedback, suffered from crosstalk or required sophisticated calibration. Here, we report a fast and high-fidelity reset scheme, avoiding the issues above without any additional chip architecture. By modulating the flux through a transmon qubit, we realize a swap between the qubit and its readout resonator that suppresses the excited state population to 0.08% $\pm$ 0.08% within 34 ns (284 ns if photon depletion of the resonator is required). Furthermore, our approach (i) can achieve effective second excited state depletion, (ii) has negligible effects on neighbouring qubits, and (iii) offers a way to entangle the qubit with an itinerant single photon, useful in quantum communication applications.

quant-ph↗