Searcharxiv⌕ Search

arXiv subjects

Peng Chen

Publications and source records attributed to Peng Chen.

At least 379 records · Page 21Linked to original sources

Approximation to the stable law by Lindeberg principle

By the Lindeberg principle, we develop in this paper an approximation to one dimensional (possibly) asymmetric $α$-stable distributions with $α\in (0,2)$ in the smooth Wasserstein distance. It is the first time that the general stable central limit theorem is proved by the Lindeberg principle, and that this theorem with $α\in (0,1]$ is proved by a new method other than Fourier analysis. Our main tools are a Taylor-like expansion and a Kolmogorov forward equation.

math.PR↗

A Versatile Software Systolic Execution Model for GPU Memory-Bound Kernels

This paper proposes a versatile high-performance execution model, inspired by systolic arrays, for memory-bound regular kernels running on CUDA-enabled GPUs. We formulate a systolic model that shifts partial sums by CUDA warp primitives for the computation. We also employ register files as a cache resource in order to operate the entire model efficiently. We demonstrate the effectiveness and versatility of the proposed model for a wide variety of stencil kernels that appear commonly in HPC, and also convolution kernels (increasingly important in deep learning workloads). Our algorithm outperforms the top reported state-of-the-art stencil implementations, including implementations with sophisticated temporal and spatial blocking techniques, on the two latest Nvidia architectures: Tesla V100 and P100. For 2D convolution of general filter sizes and shapes, our algorithm is on average 2.5x faster than Nvidia's NPP on V100 and P100 GPUs.

cs.DC↗

iFDK: A Scalable Framework for Instant High-resolution Image Reconstruction

Computed Tomography (CT) is a widely used technology that requires compute-intense algorithms for image reconstruction. We propose a novel back-projection algorithm that reduces the projection computation cost to 1/6 of the standard algorithm. We also propose an efficient implementation that takes advantage of the heterogeneity of GPU-accelerated systems by overlapping the filtering and back-projection stages on CPUs and GPUs, respectively. Finally, we propose a distributed framework for high-resolution image reconstruction on state-of-the-art GPU-accelerated supercomputers. The framework relies on an elaborate interleave of MPI collective communication steps to achieve scalable communication. Evaluation on a single Tesla V100 GPU demonstrates that our back-projection kernel performs up to 1.6x faster than the standard FDK implementation. We also demonstrate the scalability and instantaneous CT capability of the distributed framework by using up to 2,048 V100 GPUs to solve 4K and 8K problems within 30 seconds and 2 minutes, respectively (including I/O).

cs.DC↗

Optically active beams: non-reciprocal optical activity in free space induced by spin-orbital interaction of light

Optical activity, the power of a medium to rotate the polarization of a light beam, has contributed significantly to molecular structure assessments in stereochemistry, biomolecular science and crystallography. Thus far, it is commonly believed that optical activity is manifested only in the chiral media which can give rise to circular birefringence of light. Here, we experimentally demonstrate that free space can also support the implementation of non-reciprocal optical activity with Bessel beams by spin-orbital interaction. Specifically, non-diffractive optically active beams are realized, with their optical rotatory power readily controlled by simple optical elements. We show that such free-space optical activity can be exploited to form non-reciprocal optical components such as polarization rotators, isolators, and circulators. Our results may bring about new possibilities of media-independent optical activity to other transverse waves ranging from radio to optical frequencies.

physics.optics↗

BMO spaces associated to operators with generalised Poisson bounds on non-doubling manifolds with ends

Consider a non-doubling manifold with ends $M = \mathfrak{R}^{n}\sharp\, {\mathbb R}^{m}$ where $\mathfrak{R}^n=\mathbb{R}^n\times \mathbb{S}^{m-n}$ for $m> n \ge 3$. We say that an operator $L$ has a generalised Poisson kernel if $\sqrt{ L}$ generates a semigroup $e^{-t\sqrt{L}}$ whose kernel $p_t(x,y)$ has an upper bound similar to the kernel of $e^{-t\sqrtΔ}$ where $Δ$ is the Laplace-Beltrami operator on $M$. An example for operators with generalised Gaussian bounds is the Schrödinger operator $L = Δ+ V$ where $V$ is an arbitrary non-negative locally integrable potential. In this paper, our aim is to introduce the BMO space ${\rm BMO}_L(M)$ associated to operators with generalised Poisson bounds which serves as an appropriate setting for certain singular integrals with rough kernels to be bounded from $L^{\infty}(M)$ into this new ${\rm BMO}_L(M)$. On our ${\rm BMO}_L(M)$ spaces, we show that the John--Nirenberg inequality holds and we show an interpolation theorem for a holomorphic family of operators which interpolates between $L^q(M)$ and ${\rm BMO}_L(M)$. As an application, we show that the holomorphic functional calculus $m(\sqrt{L})$ is bounded from $L^{\infty}(M)$ into ${\rm BMO}_L(M)$, and bounded on $L^p(M)$ for $1 < p < \infty$.

math.AP↗

Learning Deep Representations by Mutual Information for Person Re-identification

Most existing person re-identification (ReID) methods have good feature representations to distinguish pedestrians with deep convolutional neural network (CNN) and metric learning methods. However, these works concentrate on the similarity between encoder output and ground-truth, ignoring the correlation between input and encoder output, which affects the performance of identifying different pedestrians. To address this limitation, We design a Deep InfoMax (DIM) network to maximize the mutual information (MI) between the input image and encoder output, which doesn't need any auxiliary labels. To evaluate the effectiveness of the DIM network, we propose end-to-end Global-DIM and Local-DIM models. Additionally, the DIM network provides a new solution for cross-dataset unsupervised ReID issue as it needs no extra labels. The experiments prove the superiority of MI theory on the ReID issue, which achieves the state-of-the-art results.

cs.CV↗

Distinguishing Individual Red Pandas from Their Faces

Individual identification is essential to animal behavior and ecology research and is of significant importance for protecting endangered species. Red pandas, among the world's rarest animals, are currently identified mainly by visual inspection and microelectronic chips, which are costly and inefficient. Motivated by recent advancement in computer-vision-based animal identification, in this paper, we propose an automatic framework for identifying individual red pandas based on their face images. We implement the framework by exploring well-established deep learning models with necessary adaptation for effectively dealing with red panda images. Based on a database of red panda images constructed by ourselves, we evaluate the effectiveness of the proposed automatic individual red panda identification method. The evaluation results show the promising potential of automatically recognizing individual red pandas from their faces. We are going to release our database and model in the public domain to promote the research on automatic animal identification and particularly on the technique for protecting red pandas.

cs.CV↗

Activation of nominally silent domain wall-localized phonons from GHz to THz frequencies

Ferroelectric domain walls (DWs) are nanoscale topological defects that can be easily tailored to create nanoscale devices. Their excitations, recently discovered to be responsible for DW GHz conductivity, hold promise for faster signal transmission and processing speed compared to the existing technology. Here we find that DW phonons disperse from GHz to THz frequencies, thus explaining the origin of the surprisingly broad GHz signature in DW conductivity. Puzzling activation of nominally silent DW sliding modes in BiFeO3 is traced back to DW tilting and resulting asymmetry in wall-localized phonons. The obtained phonon spectra and selection rules are used to simulate scanning impedance microscopy, emerging as a powerful probe in nanophononics. The results will guide experimental discovery of the predicted phonon branches and design of DW-based nanodevices.

cond-mat.mtrl-sci↗

Weak type $(1,1)$ bounds for Schrödinger groups

Let $L$ be a non-negative self-adjoint operator acting on $L^2(X)$ where $X$ is a space of homogeneous type with a dimension $n$. Suppose that the heat kernel of $L$ satisfies a Gaussian upper bound. It is known that the operator $(I+L)^{-s } e^{itL}$ is bounded on $L^p(X)$ for $s> n|{1/ 2}-{1/p}| $ and $ p\in (1, \infty)$ (see for example, \cite{CCO, H, Sj}). The index $s= n|{1/ 2}-{1/p}|$ was only obtained recently in \cite{CDLY, CDLY2}, and this range of $s$ is sharp since it is precisely the range known in the case when $L$ is the Laplace operator $Δ$ on $X=\mathbb R^n$ (\cite{Mi1}). In this paper we establish that for $p=1,$ the operator $(1+L)^{-n/2}e^{itL}$ is of weak type $(1, 1)$, that is, there is a constant $C$, independent of $t$ and $f$ so that \begin{eqnarray*} μ\Big(\Big\{x: \big|(I+L)^{-n/2 }e^{itL} f(x)\big|>λ\Big\}\Big) \leq Cλ^{-1}(1+|t|)^{n/2} {\|f\|_{L^1(X)} }, \ \ \ t\in{\mathbb R} \end{eqnarray*} (for $λ> 0$ when $μ(X) = \infty$ and $λ>μ(X)^{-1}\|f\|_{L^1(X)}$ when $μ(X) < \infty$). Moreover, we also show the index $n/2$ is sharp when $L$ is the Laplacian on ${\mathbb R^n}$ by providing an example. Our results are applicable to Schrödinger group for large classes of operators including elliptic operators on compact manifolds, Schrödinger operators with non-negative potentials and Laplace operators acting on Lie groups of polynomial growth or irregular non-doubling domains of Euclidean spaces.

math.AP↗

CP Symmetries as Guiding Posts: revamping tri-bi-maximal Mixing. Part II

In this follow up of arXiv:1812.04663 we analyze the generalized CP symmetries of the charged lepton mass matrix compatible with the complex version of the Tri-Bi-Maximal (TBM) lepton mixing pattern. These symmetries are used to `revamp' the simplest TBM \textit{Ansatz} in a systematic way. Our generalized patterns share some of the attractive features of the original TBM matrix and are consistent with current oscillation experiments. We also discuss their phenomenological implications both for upcoming neutrino oscillation and neutrinoless double beta decay experiments.

hep-ph↗

Giant Panda Face Recognition Using Small Dataset

Giant panda (panda) is a highly endangered animal. Significant efforts and resources have been put on panda conservation. To measure effectiveness of conservation schemes, estimating its population size in wild is an important task. The current population estimation approaches, including capture-recapture, human visual identification and collection of DNA from hair or feces, are invasive, subjective, costly or even dangerous to the workers who perform these tasks in wild. Cameras have been widely installed in the regions where pandas live. It opens a new possibility for non-invasive image based panda recognition. Panda face recognition is naturally a small dataset problem, because of the number of pandas in the world and the number of qualified images captured by the cameras in each encounter. In this paper, a panda face recognition algorithm, which includes alignment, large feature set extraction and matching is proposed and evaluated on a dataset consisting of 163 images. The experimental results are encouraging.

cs.CV↗

Comment on "Chiral Phase Transition in Charge Ordered 1T-TiSe2" and Supplementary Material on "First-order Forbidden X-ray Diffraction"

A prior report of the emergence of chirality for the (2x2x2) charge density wave (CDW) in TiSe2 has attracted much interest; the drastic symmetry breaking is highly unusual with few precedents [1]. In that study, key evidence was provided by x-ray diffraction measurements of two superlattice reflections, (1.5 1.5 0.5) and (2.5 1 0). The (2.5 1 0) reflection appeared to show an anomalously large intensity and a transition onset at ~7 K below that of the (1.5 1.5 0.5) reflection. These observations, aided by modeling, were cited as evidence for a separate chiral transition. In this Comment, we show that the prior conclusions based on x-ray diffraction are erroneous. There is just one transition, and it is achiral.

cond-mat.mtrl-sci↗

All Optical Neural Network with Nonlinear Activation Functions

Artificial neural networks (ANNs) have now been widely used for industry applications and also played more important roles in fundamental researches. Although most ANN hardware systems are electronically based, optical implementation is particularly attractive because of its intrinsic parallelism and low energy consumption. Here, we propose and demonstrate fully-functioned all optical neural networks (AONNs), in which linear operations are programmed by spatial light modulators and Fourier lenses, and optical nonlinear activation functions are realized with electromagnetically induced transparency in laser-cooled atoms. Moreover, all the errors from different optical neurons here are independent, thus the AONN could scale up to a larger system size with final error still maintaining in a similar level of a single neuron. We confirm its capability and feasibility in machine learning by successfully classifying the order and disorder phases of a typical statistic Ising model. The demonstrated AONN scheme can be used to construct various ANNs of different architectures with the intrinsic parallel computation at the speed of light.

physics.optics↗

Non-integrable stable approximation by Stein's method

We develop Stein's method for $α$-stable approximation with $α\in(0,1]$, continuing the recent line of research by Xu \cite{lihu} and Chen, Nourdin and Xu \cite{C-N-X} in the case $α\in(1,2).$ The main results include an intrinsic upper bound for the error of the approximation in a variant of Wasserstein distance that involves the characterizing differential operators for stable distributions, and an application to the generalized central limit theorem. Due to the lack of first moment for the approximating sequence in the latter result, we appeal to an additional truncation procedure and investigate fine regularity properties of the solution to Stein's equation.

math.PR↗

Sparse polynomial approximation for optimal control problems constrained by elliptic PDEs with lognormal random coefficients

In this work, we consider optimal control problems constrained by elliptic partial differential equations (PDEs) with lognormal random coefficients, which are represented by a countably infinite-dimensional random parameter with i.i.d. normal distribution. We approximate the optimal solution by a suitable truncation of its Hermite polynomial chaos expansion, which is known as a sparse polynomial approximation. Based on the convergence analysis in \cite{BachmayrCohenDeVoreEtAl2017} for elliptic PDEs with lognormal random coefficients, we establish the dimension-independent convergence rate of the sparse polynomial approximation of the optimal solution. Moreover, we present a polynomial-based sparse quadrature for the approximation of the expectation of the optimal solution and prove its dimension-independent convergence rate based on the analysis in \cite{Chen2018}. Numerical experiments demonstrate that the convergence of the sparse quadrature error is independent of the active parameter dimensions and can be much faster than that of a Monte Carlo method.

math.NA↗

Sharp spectral multipliers without semigroup framework and application to random walks

In this paper we prove spectral multiplier theorems for abstract self-adjoint operators on spaces of homogeneous type. We have two main objectives. The first one is to work outside the semigroup context. In contrast to previous works on this subject, we do not make any assumption on the semigroup. The second objective is to consider polynomial off-diagonal decay instead of exponential one. Our approach and results lead to new applications to several operators such as differential operators, pseudo-differential operators as well as Markov chains. In our general context we introduce a restriction type estimates à la Stein-Tomas. This allows us to obtain sharp spectral multiplier theorems and hence sharp Bochner-Riesz summability results. Finally, we consider the random walk on the integer lattice $\mathbf{Z}^n$ and prove sharp Bochner-Riesz summability results similar to those known for the standard Laplacian on $\mathbb{R}^n$.

math.AP↗

Predicting neutrino oscillations with "bi-large" lepton mixing matrices

We propose two schemes for the lepton mixing matrix $U = U_l^\dagger U_ν$, where $U = U_l$ refers to the charged sector, and $U_ν$ denotes the neutrino diagonalization matrix. We assume $U_ν$ to be CP conserving and its three angles to be connected with the Cabibbo angle in a simple manner. CP violation arises solely from the $U_l$, assumed to have the CKM form, $U_l\simeq V_{\rm CKM}$, suggested by unification. Oscillation parameters depend on a single parameter, leading to narrow ranges for the "solar" and "accelerator" angles $θ_{12}$ and $θ_{23}$, as well as for the CP phase, predicted as $δ_{\rm CP}\sim 1.3π$.

hep-ph↗

Analyzing a single-laser repumping scheme for efficient loading of a strontium magneto-optical trap

We demonstrate enhanced loading of strontium atoms into a magneto-optical trap using a repumping scheme from the metastable state via the doubly-excited state $5\mathrm{s}5\mathrm{p}\,^3\mathrm{P}_2 \rightarrow 5\mathrm{p}^2\,^3\mathrm{P}_2$ at $481~\mathrm{nm}$. The number of trapped atoms is increased by an order of magnitude. The frequency and intensity dependence of the atom number enhancement, with respect to the non-repumping case, is well reproduced by a simple rate equation model, which also describes single-laser repumping schemes reported previously. The repumping scheme is limited by a weak additional loss channel into the long-lived $5\mathrm{s}5\mathrm{p}\,^3\mathrm{P}_0$ state. For low repumping intensities, the signature of a halo formed by magnetically trapped atoms in the metastable state is found.

physics.atom-ph↗