arXiv subjects

Yijun Wang

Publications and source records attributed to Yijun Wang.

99 records · Page 6Linked to original sources

arXiv2020

BETA: A Large Benchmark Database Toward SSVEP-BCI Application

Brain-computer interface (BCI) provides an alternative means to communicate and it has sparked growing interest in the past two decades. Specifically, for Steady-State Visual Evoked Potential based BCI, marked improvement has been made in the frequency recognition method and data sharing. However, the number of pubic database is still limited in this field. Therefore, we present a \textbf{BE}nchmark database \textbf{T}owards BCI \textbf{A}pplication (BETA) in the study. The BETA database is composed of 64-channel Electroencephalogram (EEG) data from 70 subjects performing a 40-target cued-spelling task. The design and acquisition of BETA is in pursuit of meeting the demand from real-world applications and it can be used as a test-bed for these scenarios. We validate the database by a series of analysis and conduct the classification analysis of eleven frequency recognition methods on BETA. We recommend to use the metric of wide-band SNR and BCI quotient to characterize the SSVEP at the single-trial and population level, respectively. The BETA database can be downloaded from the website http://bci.med.tsinghua.edu.cn/download.html.

eess.SP↗

arXiv2019

Ensemble Pruning based on Objection Maximization with a General Distributed Framework

Ensemble pruning, selecting a subset of individual learners from an original ensemble, alleviates the deficiencies of ensemble learning on the cost of time and space. Accuracy and diversity serve as two crucial factors while they usually conflict with each other. To balance both of them, we formalize the ensemble pruning problem as an objection maximization problem based on information entropy. Then we propose an ensemble pruning method including a centralized version and a distributed version, in which the latter is to speed up the former. At last, we extract a general distributed framework for ensemble pruning, which can be widely suitable for most of the existing ensemble pruning methods and achieve less time consuming without much accuracy degradation. Experimental results validate the efficiency of our framework and methods, particularly concerning a remarkable improvement of the execution speed, accompanied by gratifying accuracy performance.

cs.LG↗

arXiv2019

A narrow-band sodium-resonant fiber-coupled single photon source

Quantum technology requires the creation and control over single photons as an important resource. We present a single photon source based on a single molecule which is attached to the end-facet of an optical fiber. To realize a narrow linewidth, the system is cooled down to liquid-helium temperatures. The molecule is optically excited and its fluorescence is collected through the fiber. We have recorded an excitation spectrum, a saturation curve and analyzed the contributions of Raman background fluorescence. This presents to date the crucial limit for the introduced device. The single photon nature is proven by an anti-bunched auto-correlation recording, which also shows coherent Rabi oscillations.

quant-ph↗

arXiv2019

X-ray spectral variations of synchrotron peak in BL Lacs

The spectral energy distribution of blazars around the synchrotron peak can be well described by the log-parabolic model that has three parameters: peak energy ($E_\textrm{p}$), peak luminosity ($L_\textrm{p}$) and the curvature parameter ($b$). It has been suggested that $E_\textrm{p}$ shows relations with $L_\textrm{p}$ and $b$ in several sources, which can be used to constrain the physical properties of the emitting region and/or acceleration processes of the emitting particles. We systematically study the $E_\textrm{p}$-$L_\textrm{p}$ and $E_\textrm{p}$-(1$/b$) relations for 14 BL Lac objects using the 3-25~keV $RXTE$/PCA and 0.3-10~keV $Swift$/XRT data. Most objects (9/14) exhibit positive $E_\textrm{p}$-$L_\textrm{p}$ correlations, three sources show no correlation, and two sources display negative correlations. In addition, most targets (7/14) present no correlation between $E_\textrm{p}$ and 1$/b$, five sources pose negative correlations, and two sources demonstrate positive correlations. 1ES~1959+650 displays two different $E_\textrm{p}$-$L_\textrm{p}$ relations in 2002 and 2016. We also analyze $E_\textrm{p}$-$L_\textrm{p}$ and $E_\textrm{p}$-(1$/b$) relations during flares lasting for several days. The $E_\textrm{p}$-$L_\textrm{p}$ relation does not exhibit significant differences between flares, while the $E_\textrm{p}$-(1$/b$) relation varies from flare to flare. For the total sample, when $L_\textrm{p}$ < $\textrm{10}^\textrm{45}\ \textrm{erg}\ \textrm{s}^\textrm{-1}$, there seems to be a positive $E_\textrm{p}$-$L_\textrm{p}$ correlation. $L_\textrm{p}$ and the slope of $E_\textrm{p}$-$L_\textrm{p}$ relation present an anti-correlation, which indicates that the causes of spectral variations might be different between luminous and faint sources. $E_\textrm{p}$ shows a positive correlation with the black hole mass. We discuss the implications of these results.

astro-ph.GA↗

arXiv2019

Learning to Transfer: Unsupervised Meta Domain Translation

Unsupervised domain translation has recently achieved impressive performance with Generative Adversarial Network (GAN) and sufficient (unpaired) training data. However, existing domain translation frameworks form in a disposable way where the learning experiences are ignored and the obtained model cannot be adapted to a new coming domain. In this work, we take on unsupervised domain translation problems from a meta-learning perspective. We propose a model called Meta-Translation GAN (MT-GAN) to find good initialization of translation models. In the meta-training procedure, MT-GAN is explicitly trained with a primary translation task and a synthesized dual translation task. A cycle-consistency meta-optimization objective is designed to ensure the generalization ability. We demonstrate effectiveness of our model on ten diverse two-domain translation tasks and multiple face identity translation tasks. We show that our proposed approach significantly outperforms the existing domain translation methods when each domain contains no more than ten training samples.

cs.CV↗

arXiv2019

Image-to-Image Translation with Multi-Path Consistency Regularization

Image translation across different domains has attracted much attention in both machine learning and computer vision communities. Taking the translation from source domain $\mathcal{D}_s$ to target domain $\mathcal{D}_t$ as an example, existing algorithms mainly rely on two kinds of loss for training: One is the discrimination loss, which is used to differentiate images generated by the models and natural images; the other is the reconstruction loss, which measures the difference between an original image and the reconstructed version through $\mathcal{D}_s\to\mathcal{D}_t\to\mathcal{D}_s$ translation. In this work, we introduce a new kind of loss, multi-path consistency loss, which evaluates the differences between direct translation $\mathcal{D}_s\to\mathcal{D}_t$ and indirect translation $\mathcal{D}_s\to\mathcal{D}_a\to\mathcal{D}_t$ with $\mathcal{D}_a$ as an auxiliary domain, to regularize training. For multi-domain translation (at least, three) which focuses on building translation models between any two domains, at each training iteration, we randomly select three domains, set them respectively as the source, auxiliary and target domains, build the multi-path consistency loss and optimize the network. For two-domain translation, we need to introduce an additional auxiliary domain and construct the multi-path consistency loss. We conduct various experiments to demonstrate the effectiveness of our proposed methods, including face-to-face translation, paint-to-photo translation, and de-raining/de-noising translation.

cs.CV↗

arXiv2018

Systematic investigation of X-ray spectral variability of TeV blazars during flares in the $RXTE$ era

Utilizing all the 16-year $RXTE$ observations, we analyze the X-ray spectra of 32 TeV blazars, and perform a systematic investigation of X-ray spectral variability for the 5 brightest sources during their major flares that lasted several days. We obtain photon spectral index ($α$), flux and synchrotron radiation peak energy ($E_{\rm{p}}$) from empirical spectral fitting, and electron spectral index ($p$) from theoretical synchrotron radiation modeling. We find that both $α$ and $p$ generally display a harder-when-brighter trend, confirming the results of many previous works. Furthermore, we confirm and strengthen the result that $p$ must vary in order to explain the observed X-ray spectral variability during flares, which would have useful implications for interpreting the associated higher-energy spectral variability. We see apparent electron spectral hysteresis in many but not all $p$-flux plots that takes a form of "loop" or oblique"8". We obtain a tight $p$-hardness ratio (HR) relation and a tighter $p$-$α$ relation using spectra of flaring periods, both of which are also applicable to stacked data of quiescent periods. We demonstrate that these two empirical relations can be used efficiently to estimate $p$ from HR or $α$ that is readily achieved. Finally, we find that, when considering TeV blazars as a whole, $α$ and X-ray luminosity are positively correlated, $E_{\rm{p}}$ is negatively correlated with $p$ and $α$, and $E_{\rm{p}}$ is positively correlated with HR; all these correlations are in line with the blazar sequence. However, after correcting for the Doppler boosting effect, $α$ and intrinsic X-ray luminosity follow an anti-correlation.

astro-ph.GA↗

arXiv2018

Dual-phase-modulated plug-and-play measurement-device-independent continuous-variable quantum key distribution

We suggest an improved plug-and-play measurement-device-independent (MDI) continuous-variable quantum key distribution (CVQKD) via the dual-phase modulation (DPM), aiming to solve an implementation problem with no extra performance penalty. The synchronous loophole of different lasers from Alice and Bob can be elegantly eliminated in the plug-and-play configuration, which gives birth to the convenient implementation when comparing to the Gaussian-modulated coherent-state protocol. While the local oscillator (LO) can be locally generated by the trusted part Charlie, the LO-aimed attacks can be accurately detected in the data post-processing. We derive the security bounds of the DPM-based MDI-CVQKD against optimal Gaussian collective attacks. Taking the finite-size effect into account, the secret key rate can be increased due to the fact that almost all raw keys of the MDI-CVQKD system can be fully exploited for the final secret key generation without sacrificing raw keys in parameter estimation. Moreover, we give an experimental concept of the proposed scheme which can be deemed guideline for final implementation.

quant-ph↗

arXiv2013

Classifying Single-Trial EEG during Motor Imagery with a Small Training Set

Before the operation of a motor imagery based brain-computer interface (BCI) adopting machine learning techniques, a cumbersome training procedure is unavoidable. The development of a practical BCI posed the challenge of classifying single-trial EEG with a small training set. In this letter, we addressed this problem by employing a series of signal processing and machine learning approaches to alleviate overfitting and obtained test accuracy similar to training accuracy on the datasets from BCI Competition III and our own experiments.

cs.LG↗