arXiv subjects

Xin Yu

Publications and source records attributed to Xin Yu.

At least 271 records · Page 15Linked to original sources

arXiv2018

Identity-preserving Face Recovery from Portraits

Recovering the latent photorealistic faces from their artistic portraits aids human perception and facial analysis. However, a recovery process that can preserve identity is challenging because the fine details of real faces can be distorted or lost in stylized images. In this paper, we present a new Identity-preserving Face Recovery from Portraits (IFRP) to recover latent photorealistic faces from unaligned stylized portraits. Our IFRP method consists of two components: Style Removal Network (SRN) and Discriminative Network (DN). The SRN is designed to transfer feature maps of stylized images to the feature maps of the corresponding photorealistic faces. By embedding spatial transformer networks into the SRN, our method can compensate for misalignments of stylized faces automatically and output aligned realistic face images. The role of the DN is to enforce recovered faces to be similar to authentic faces. To ensure the identity preservation, we promote the recovered and ground-truth faces to share similar visual features via a distance measure which compares features of recovered and ground-truth faces extracted from a pre-trained VGG network. We evaluate our method on a large-scale synthesized dataset of real and stylized face pairs and attain state of the art results. In addition, our method can recover photorealistic faces from previously unseen stylized portraits, original paintings and human-drawn sketches.

cs.CV↗

arXiv2018

Face Destylization

Numerous style transfer methods which produce artistic styles of portraits have been proposed to date. However, the inverse problem of converting the stylized portraits back into realistic faces is yet to be investigated thoroughly. Reverting an artistic portrait to its original photo-realistic face image has potential to facilitate human perception and identity analysis. In this paper, we propose a novel Face Destylization Neural Network (FDNN) to restore the latent photo-realistic faces from the stylized ones. We develop a Style Removal Network composed of convolutional, fully-connected and deconvolutional layers. The convolutional layers are designed to extract facial components from stylized face images. Consecutively, the fully-connected layer transfers the extracted feature maps of stylized images into the corresponding feature maps of real faces and the deconvolutional layers generate real faces from the transferred feature maps. To enforce the destylized faces to be similar to authentic face images, we employ a discriminative network, which consists of convolutional and fully connected layers. We demonstrate the effectiveness of our network by conducting experiments on an extensive set of synthetic images. Furthermore, we illustrate our network can recover faces from stylized portraits and real paintings for which the stylized data was unavailable during the training phase.

cs.CV↗

arXiv2017

Solitons and breathers for nonisospectral mKdV equation with Darboux transformation

Under investigation in this paper is the nonisospectral and variable coefficients modified Kortweg-de Vries (vc-mKdV) equation, which manifests in diverse areas of physics such as fluid dynamics, ion acoustic solitons and plasma mechanics. With the degrees of restriction reduced, a simplified constraint is introduced, under which the vc-mKdV equation is an integrable system and the spectral flow is time-varying. The Darboux transformation for such equation is constructed, which gives rise to the generation of variable kinds of solutions including the double-breather coherent structure, periodical soliton-breather and localized solitons and breathers. In addition, the effect of variable coefficients and initial phases is discussed in terms of the soliton amplitude, polarity, velocity and width, which might provide feasible soliton management with certain conditions taken into account.

nlin.PS↗

arXiv2017

Darboux Transformation for the Nonisospectral and Variable-coefficient KdV Equation

With the nonuniform media taken into account, the nonisospectral and variable-coefficient Korteweg-de Vries equation, which describes various physical situations such as fluid dynamics and plasma, is under investigation in this paper. With appropriate selection of wave functions, the Darboux transformation is constructed, by which the multi-soliton solutions are derived and graphs are presented. The spectral parameters, coefficients and initial phase are discussed analytically and numerically to demonstrate their respective effect on the soliton dynamics, which plays a role in achieving the feasible soliton management with explicit conditions taken into account.

nlin.PS↗

arXiv2017

Spacial inhomogeneity and nonlinear tunneling for the forced KdV equation

A variable-coefficient forced Korteweg-de Vries equation with spacial inhomogeneity is investigated in this paper. Under constraints, this equation is transformed into its bilinear form, and multi-soliton solutions are derived. Effects of spacial inhomogeneity for soliton velocity, width and background are discussed. Nonlinear tunneling for this equation is presented, where the soliton amplitude can be amplified or compressed. Our results might be useful for the relevant problems in fluids and plasmas.

nlin.PS↗

arXiv2014

Semileptonic decays $B_c^+\to D^{(*)}_{(s)}(l^+ν,l^+l^-,ν\barν)$ in the perturbative QCD approach

In this paper we study the semileptonic decays of $B_c^+\to D^{(*)}_{(s)}(l^+ν_l,l^+l^-,ν\barν)$ (here $l$ stands for $e$, $μ$, or $τ$). After evaluating the $B_c^+ \to (D_{(s)},D^*_{(s)})$ transition form factors $F_{0,+,T}(q^2)$ and $V(q^2), A_{0,1,2}(q^2), T_{1,2,3}(q^2)$ by employing the perturbative QCD factorization approach, we calculate the branching ratios for all these semileptonic decays. Our predictions for the values of the $B_c^+ \to D_{(s)}$ and $B_c^+ \to D^*_{(s)}$ transition form factors are consistent with those obtained by using other methods. The branching ratios of the decay modes with $\barνν$ are almost an order of magnitude larger than the corresponding decays with $l^+l^-$ after the summation over the three neutrino generations. The branching ratios for the decays with $b\to d$ transitions are much smaller than those decays with the $b\to s$ transitions, due to the Cabibbo-Kobayashi-Maskawa suppression. We define ratios $R_D$ and $R_{D^*}$ for the branching ratios with the $τ$ lepton versus $μ$, $e$ lepton final states to cancel the uncertainties of the form factors, which could possibly be tested in the near future.

hep-ph↗

arXiv2014

The NLO twist-3 contributions to $B \to π$ form factors in $k_{T}$ factorization

In this paper, we calculate the next-to-leading-order (NLO) twist-3 contribution to the form factors of $B \to π$ transitions by employing the $k_{T}$ factorization theorem. All the infrared divergences regulated by the logarithms $\ln(k_{iT}^{2})$ cancel between those from the quark diagrams and from the effective diagrams for the initial $B$ meson wave function and the final pion meson wave function. An infrared finite NLO hard kernel is therefore obtained, which confirms the application of the $k_{T}$ factorization theorem to $B$ meson semileptonic decays at twist-3 level. From our analytical and numerical evaluations, we find that the NLO twist-3 contributions to the form factors $f^{+,0}(q^2)$ of $B \to π$ transition are similar in size, but have an opposite sign with the NLO twist-2 contribution, which leads to a large cancelation between these two NLO parts. For the case of $f^+(0)$, for example, the $24\%$ NLO twist-2 enhancement to the full LO prediction is largely canceled by the negative ( about $-17\%$ ) NLO twist-3 contribution, leaving a small and stable $7\%$ enhancement to the full LO prediction in the whole range of $0\leq q^2\leq 12$ GeV$^2$. At the full NLO level, the perturbative QCD prediction is $F^{B \to π}(0)=0.269^{+0.054}_{-0.050}$. We also studied the possible effects on the pQCD predictions when different sets of the B meson and pion distribution amplitudes are used in the numerical evaluation.

hep-ph↗

arXiv2014

Global existence of null-form wave equations on small asymptotically Euclidean manifolds

We prove the global existence of the small solutions to the Cauchy problem for quasilinear wave equations satisfying the null condition on $(R^3, g)$, where the metric $g$ is a small perturbation of the flat metric and approaches the Euclidean metric like $(1+|x|)^{-a}$ with $a>1$. Global and almost global existence for systems without the null condition are also discussed for certain small time-dependent perturbations of the flat metric in the appendix.

math.AP↗

arXiv2014

Perturbative QCD study of $B_s$ decays to a pseudoscalar meson and a tensor meson

We study two-body hadronic $B_s\to PT$ decays, with $P (T)$ being a light pseudoscalar (tensor) meson, in the perturbative QCD approach. The CP-averaged branching ratios and the direct CP asymmetries of the $ΔS=0$ modes are predicted, where $ΔS$ is the difference between the strange numbers of final and initial states. We also define and calculate experimental observables for the $ΔS=1$ modes under the $B_s^0-\bar{B}_s^0$ mixing, including CP averaged branching ratios, time-integrated CP asymmetries, and the CP observables $C_{f}$, $D_{f}$ and $S_{f}$. Results are compared to the $B_s\to PV$ ones in the literature, and to the $B\to PT$ ones, which indicate considerable U-spin symmetry breaking. Our work provides theoretical predictions for the $B_s\to PT$ decays for the first time, some of which will be potentially measurable at future experiments.

hep-ph↗

arXiv2013

Time-dependent CP-violations of B(Bs) decays in the perturbative QCD approach

We study the decay modes of B_{s}^{0}(\bar{B}_{s}^{0})-->D_{s}^{\pm} K^{\mp}, B_{s}^{0}(\bar{B}_{s}^{0})-->D^{\pm} π^{\mp} and B^{0}(\bar{B}^{0})-->D^{\pm} π^{\mp} in the perterbative QCD approach based on k_T factorization, including the branching ratios and CP violation parameters which provide a clear way to extract the Cabibbo-Kobayashi-Maskawa angle γ. Our results of branching ratios of B_{s}^{0}(\bar{B}_{s}^{0})-->D_{s}^{\pm} K^{\mp} and B^{0}(\bar{B}^{0})--> D^{\pm} π^{\mp} and the CP asymmetry of B^{0}(\bar{B}^{0})--> D^{\pm} π^{\mp} agree well with the experimental data. We also give the predictions of the other observables, which provide some guidance for experiments in the future, especially for LHCb experiment.

hep-ph↗

arXiv2012

Charmed B(B_{s}) decays involving a light tensor meson in PQCD approach

We study the $B(B_{s})\rightarrow D_{(s)}(\bar{D}_{(s)})\,T$ and $D_{(s)}^{\ast}(\bar{D}_{(s)}^{*})\,T$ decays in perturbative QCD approach, where T denotes a light tensor meson. In addition to the branching ratios, we also give predictions to the polarization fractions for the $D^{*}_{(s)}(\bar{D}_{(s)}^{*})T$ decays. For those decays with a tensor meson emitted, the factorizable emission diagrams do not contribute because of lorentz covariance considerations. In order to give the predictions to these decay channels, it is necessary to go beyond the naive factorization to calculate the nonfactorizable and annihilation diagrams.

hep-ph↗

arXiv2012

The $B_c\rightarrow D^{(*)}T$ decays in perturbative QCD approach

In this work, we investigate those $B_{c}\rightarrow D^{(*)}T$ decays in perturbative QCD approach, based on $k_T$ factorization, where T denotes a light tensor meson. For all decays considered in this work, there are no contributions from factorizable emission diagrams because the emitted meson is the tensor meson. We find that the annihilation amplitudes are dominant in these decays due to the large Cabibbo-Kobayashi-Maskawa elements, which are only calculable in the pQCD approach. The numerical results show that the predictions for the branching ratios of most decays are in the order of $10^{-6}$ or even bigger, which can be observed in the ongoing experiments. We also predict large percentage of transverse polarizations in those W annihilation diagram dominant $B_{c} \rightarrow D^{*}T$ decay channels.

hep-ph↗

arXiv2012

Charmed Scalar Meson Production in $B$ Decays

The study on the charmed scalar meson spectroscopy has become a hot topic both experimentally and theoretically. The $B_{(s)}$ decays provide an ideal place to study their property. We employ the $B$-meson light-cone sum rules to compute the $\bar B_s^0\to D_s^{*+}(2317)$ and $B^-\to D_0^{*0}(2400)$ transition form factors at large recoil, assuming $D_s^{*+}(2317)$ and $D_0^{*0}(2400)$ being scalar quark-anti-quark states. The results are extrapolated to the whole momentum region with the help of HQET. Considering large uncertainties, our results can be consistent with the previous studies, while the power corrections should be large. We also estimate the semi-leptonic decays $\bar B_s^0\to D_s^{*+}(2317)l \barν_l$ and $B^-\to D_0^{*0}(2400)l \barν_l$. The branching fraction of the semi-leptonic $\bar B_s^0\to D_s^{*+}(2317)l \barν_l$ decay is around $6\times 10^{-3} $ for light leptons and $0.8\times 10^{-3} $ for tau final state. The predicted branching ration of $B^-\to D_0^{*0}(2400)l \barν_l$ is slightly larger than $\bar B_s^0\to D_s^{*+}(2317)l \barν_l$, and we hope the future data in LHCb can test these results.

hep-ph↗

arXiv2012

The $B(B_{s})\rightarrow D_{(s)}(\bar{D}_{(s)}) T$ and $D_{(s)}^{}(\bar{D}_{(s)}^{})T$ decays in perturbative QCD approach

In perturbative QCD approach, we investigate the $B(B_{s})\rightarrow D_{(s)}(\bar{D}_{(s)})\,T$ and $D_{(s)}^{*}(\bar{D}_{(s)}^{*})\,T$ decays, which include the Cabibbo-Kobayashi-Maskawa (CKM) favored decays and the Cabibbo-Kobayashi-Maskawa-suppressed decays, where T denotes a light tensor meson. From our calculation, we find that the nonfactorizable emission diagrams and the annihilation type diagrams are important, especially for those color suppressed channels. For those decays with a tensor meson emitted, the factorizable emission diagrams vanish owing to the fact that a tensor meson can not be produced through the local (V-A) or tensor current. The numerical results show that the predictions for the branching ratios of considered charmed B decays are in the range of $10^{-4}$ to $10^{-6}$ for those CKM-favored decays (governed by $|V_{cb}|$) and in the range of $10^{-5}$ to $10^{-8}$ for those CKM-suppressed decays (governed by $|V_{ub}|$). We also predict large transverse polarization contributions in many of the $B(B_{s}) \rightarrow D_{(s)}^{\ast}(\bar{D}_{(s)}^{*})T$ decay channels.

hep-ph↗

arXiv2012

Recent works on the Strauss conjecture

In this review paper, we summarize the current state-of-art on the Strauss conjecture with nontrapping obstacles. Among others, three essential estimates are emphasized and presented: Morawetz-KSS estimates (also known as local energy estimates), weighted Strichartz estimates and generalized Strichartz estimates.

math.AP↗

arXiv2012

Nonleptonic two-body charmless B decays involving a tensor meson in the Perturbative QCD Approach

Two-body charmless hadronic B decays involving a light tensor meson in the final states are studied in the perturbative QCD approach based on $k_T$ factorization. From our calculations, we find that the decay branching ratios for color allowed tree-dominated decays $B\to a_{2}^{0}π^{+}$ and $B\to a_{2}^{-}π^{+}$ modes are of order $10^{-6}$ and $10^{-5}$, respectively. While other color suppressed tree-dominated decays have very small branching ratios. In general, the branching ratios of most decays are in the range of $10^{-5}$ to $10^{-8}$, which are bigger by one or two orders of magnitude than those predictions obtained in Isgur-Scora-Grinstein-Wise II model and in the covariant light-front approach, but consistent with the recent experimental measurements and the QCD factorization calculations. Since the decays with a tensor meson emitted from vacuum are prohibited in naive factorization, the contributions of nonfactorizable and annihilation diagrams are very important to these decays, which are calculable in our perturbative QCD approach. We also give predictions to the direct CP asymmetries, some of which are large enough for the future experiments to measure. Because we considered the mixing between $f_{2}$ and $f_{2}'$, the decay rates are enhanced significantly for some decays involving $f_{2}^{\prime}$ meson, even with a small mixing angle.

hep-ph↗

arXiv2011

Circuit Modeling of Tunneling Real-Space Transfer Transistors: Toward Terahertz Frequency Operation

High frequency operation of tunneling real-space transfer transistor (TRSTT) in the negative differential resistance (NDR) regime is assessed by calculating the device common source unity current gain frequency (fT) range with a small signal equivalent circuit model including tunneling. Our circuit model is based on an In0.2Ga0.8As and delta-doped GaAs dual channel structure with various gate lengths. The calculated TRSTT fT agrees very well with experimental data, limiting factor being the resistance of the delta-doped GaAs layer. By optimizing the gate dimensions and channel materials, we find fT in the NDR region approaches terahertz range, which anticipates potential use of TRSTT as terahertz sources.

cond-mat.mes-hall↗

arXiv2011

Generalized and weighted Strichartz estimates

In this paper, we explore the relations between different kinds of Strichartz estimates and give new estimates in Euclidean space $\mathbb{R}^n$. In particular, we prove the generalized and weighted Strichartz estimates for a large class of dispersive operators including the Schrödinger and wave equation. As a sample application of these new estimates, we are able to prove the Strauss conjecture with low regularity for dimension 2 and 3.

math.AP↗