arXiv subjects

Mei Yu

Publications and source records attributed to Mei Yu.

24 records · Page 2Linked to original sources

arXiv2020

Attention-based Residual Speech Portrait Model for Speech to Face Generation

Given a speaker's speech, it is interesting to see if it is possible to generate this speaker's face. One main challenge in this task is to alleviate the natural mismatch between face and speech. To this end, in this paper, we propose a novel Attention-based Residual Speech Portrait Model (AR-SPM) by introducing the ideal of the residual into a hybrid encoder-decoder architecture, where face prior features are merged with the output of speech encoder to form the final face feature. In particular, we innovatively establish a tri-item loss function, which is a weighted linear combination of the L2-norm, L1-norm and negative cosine loss, to train our model by comparing the final face feature and true face feature. Evaluation on AVSpeech dataset shows that our proposed model accelerates the convergence of training, outperforms the state-of-the-art in terms of quality of the generated face, and achieves superior recognition accuracy of gender and age compared with the ground truth.

cs.CV↗

arXiv2020

Magnetostrictively induced stationary entanglement between two microwave fields

We present a scheme to entangle two microwave fields by using the nonlinear magnetostrictive interaction in a ferrimagnet. The magnetostrictive interaction enables the coupling between a magnon mode (spin wave) and a mechanical mode in the ferrimagnet, and the magnon mode simultaneously couples to two microwave cavity fields via the magnetic dipole interaction. The magnon-phonon coupling is enhanced by directly driving the ferrimagnet with a strong red-detuned microwave field, and the driving photons are scattered onto two sidebands induced by the mechanical motion. We show that two cavity fields can be prepared in a stationary entangled state if they are respectively resonant with two mechanical sidebands. The present scheme illustrates a new mechanism for creating entangled states of optical fields, and enables potential applications in quantum information science and quantum tasks that require entangled microwave fields.

quant-ph↗

arXiv2019

Macroscopic entanglement of two magnon modes via quantum correlated microwave fields

We present a scheme to entangle two magnon modes in two macroscopic yttrium-iron-garnet spheres. The two spheres are placed inside two microwave cavities, which are driven by a two-mode squeezed microwave field. By using the linear state-swap interaction between the cavity and the magnon mode in each cavity, the quantum correlation of the two driving fields is with high efficiency transferred to the two magnon modes. Considerable entanglement could be achieved under experimentally achievable conditions $g \gg κ_a \gg κ_m$, where $g$ is the cavity-magnon coupling rate and $κ_a$, $κ_m$ are the decay rates of the cavity and magnon modes, respectively. The entanglement is in the steady state and robust against temperature, surviving up to hundreds of milliKelvin with experimentally accessible two-mode squeezed source.

quant-ph↗

arXiv2018

Scene Learning: Deep Convolutional Networks For Wind Power Prediction by Embedding Turbines into Grid Space

Wind power prediction is of vital importance in wind power utilization. There have been a lot of researches based on the time series of the wind power or speed, but In fact, these time series cannot express the temporal and spatial changes of wind, which fundamentally hinders the advance of wind power prediction. In this paper, a new kind of feature that can describe the process of temporal and spatial variation is proposed, namely, Spatio-Temporal Features. We first map the data collected at each moment from the wind turbine to the plane to form the state map, namely, the scene, according to the relative positions. The scene time series over a period of time is a multi-channel image, i.e. the Spatio-Temporal Features. Based on the Spatio-Temporal Features, the deep convolutional network is applied to predict the wind power, achieving a far better accuracy than the existing methods. Compared with the starge-of-the-art method, the mean-square error (MSE) in our method is reduced by 49.83%, and the average time cost for training models can be shortened by a factor of more than 150.

cs.LG↗

arXiv2017

A Liouville Theorem for a Class of Fractional Systems in $\mathbb{R}^n_+$

Let $0<α,β<2$ be any real number. In this paper, we investigate the following semilinear system involving the fractional Laplacian \begin{equation*} \left\{\begin{array}{lll} (-\lap)^{α/2} u(x)=f(v(x)), & (-\lap)^{β/2} v(x)=g(u(x)), & \qquad x\in\mathbb{R}^n_+, u,v\geq0, & \qquad x\in\mathbb{R}^n\setminus\mathbb{R}^n_+. \end{array}\right. \end{equation*} Applying a direct method of moving planes for the fractional Laplacian, without any decay assumption on the solutions at infinity, we prove Liouville theorems of nonnegative solutions under some natural conditions on $f$ and $g$.

math.AP↗

arXiv2016

Solutions of Fully Nonlinear Nonlocal Systems

In this paper we consider the system involving fully nonlinear nonlocal operators: $$ \left\{ \begin{array}{ll} F_α(u(x)) = C_{n,α} PV \int_{{R}^n} \frac{G(u(x)-u(y))}{|x-y|^{n+α}} dy=f(v(x)), F_β(v(x)) = C_{n,β} PV \int_{{R}^n} \frac{G(v(x)-v(y))}{|x-y|^{n+β}} dy=g(u(x)). \end{array} \right. $$ A \textit{narrow region principle} and a \textit{decay at infinity} for the system for carrying on the method of moving planes are established. Then we prove the radial symmetry and monotonicity for positive solutions to the nonlinear system in the whole space. Non-existence of positive solutions to the nonlinear system on a half space is proved.

math.AP↗