arXiv subjects

Bo Zhang

Publications and source records attributed to Bo Zhang.

At least 649 records · Page 36Linked to original sources

arXiv2021

Convergence of the uniaxial PML method for time-domain electromagnetic scattering problems

In this paper, we propose and study the uniaxial perfectly matched layer (PML) method for three-dimensional time-domain electromagnetic scattering problems, which has a great advantage over the spherical one in dealing with problems involving anisotropic scatterers. The truncated uniaxial PML problem is proved to be well-posed and stable, based on the Laplace transform technique and the energy method. Moreover, the $L^2$-norm and $L^{\infty}$-norm error estimates in time are given between the solutions of the original scattering problem and the truncated PML problem, leading to the exponential convergence of the time-domain uniaxial PML method in terms of the thickness and absorbing parameters of the PML layer. The proof depends on the error analysis between the EtM operators for the original scattering problem and the truncated PML problem, which is different from our previous work (SIAM J. Numer. Anal. 58(3) (2020), 1918-1940).

math.AP↗

arXiv2021

Efficient Compressed Sensing Based Image Coding by Using Gray Transformation

In recent years, compressed sensing (CS) based image coding has become a hot topic in image processing field. However, since the bit depth required for encoding each CS sample is too large, the compression performance of this paradigm is unattractive. To address this issue, a novel CS-based image coding system by using gray transformation is proposed. In the proposed system, we use a gray transformation to preprocess the original image firstly and then use CS to sample the transformed image. Since gray transformation makes the probability distribution of CS samples centralized, the bit depth required for encoding each CS sample is reduced significantly. Consequently, the proposed system can considerably improve the compression performance of CS-based image coding. Simulation results show that the proposed system outperforms the traditional one without using gray transformation in terms of compression performance.

cs.MM↗

arXiv2021

Stellar parametrization of LAMOST M dwarf stars

M dwarf stars are the most common stars in the Galaxy, dominating the population of the Galaxy by numbers at faint magnitudes. Precise and accurate stellar parameters for M dwarfs are of crucial importance for many studies. However, the atmospheric parameters of M dwarf stars are difficult to be determined. In this paper, we present a catalog of the spectroscopic stellar parameters ($T_{eff}$ and [M/H]) of $\sim$ 300,000 M dwarf stars observed by both LAMOST and Gaia using Stellar Label Machine (SLAM). We train a SLAM model using LAMOST spectra with APOGEE Data Release 16 (DR16) labels with $2800 \lt T_{eff} \lt 4500$K and $-2 \lt [M/H] \lt 0.5$ dex. The SLAM $T_{eff}$ is in agreement to within $\sim 50$K compared to the previous study determined by APOGEE observation, and SLAM [M/H] agree within 0.12 dex compared to the APOGEE observation. We also set up a SLAM model trained by BT-Settl atmospheric model, with random uncertainties (in cross-validation) to 60K and agree within $\sim 90$K compared to previous study.

astro-ph.SR↗

arXiv2021

User-Level Privacy-Preserving Federated Learning: Analysis and Performance Optimization

Federated learning (FL), as a type of collaborative machine learning framework, is capable of preserving private data from mobile terminals (MTs) while training the data into useful models. Nevertheless, from a viewpoint of information theory, it is still possible for a curious server to infer private information from the shared models uploaded by MTs. To address this problem, we first make use of the concept of local differential privacy (LDP), and propose a user-level differential privacy (UDP) algorithm by adding artificial noise to the shared models before uploading them to servers. According to our analysis, the UDP framework can realize $(ε_{i}, δ_{i})$-LDP for the $i$-th MT with adjustable privacy protection levels by varying the variances of the artificial noise processes. We then derive a theoretical convergence upper-bound for the UDP algorithm. It reveals that there exists an optimal number of communication rounds to achieve the best learning performance. More importantly, we propose a communication rounds discounting (CRD) method. Compared with the heuristic search method, the proposed CRD method can achieve a much better trade-off between the computational complexity of searching and the convergence performance. Extensive experiments indicate that our UDP algorithm using the proposed CRD method can effectively improve both the training efficiency and model quality for the given privacy protection levels.

cs.LG↗

arXiv2021

Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation

Self-training is a competitive approach in domain adaptive segmentation, which trains the network with the pseudo labels on the target domain. However inevitably, the pseudo labels are noisy and the target features are dispersed due to the discrepancy between source and target domains. In this paper, we rely on representative prototypes, the feature centroids of classes, to address the two issues for unsupervised domain adaptation. In particular, we take one step further and exploit the feature distances from prototypes that provide richer information than mere prototypes. Specifically, we use it to estimate the likelihood of pseudo labels to facilitate online correction in the course of training. Meanwhile, we align the prototypical assignments based on relative feature distances for two different views of the same target, producing a more compact target feature space. Moreover, we find that distilling the already learned knowledge to a self-supervised pretrained model further boosts the performance. Our method shows tremendous performance advantage over state-of-the-art methods. We will make the code publicly available.

cs.CV↗

arXiv2021

Deep Sketch-guided Cartoon Video Inbetweening

We propose a novel framework to produce cartoon videos by fetching the color information from two input keyframes while following the animated motion guided by a user sketch. The key idea of the proposed approach is to estimate the dense cross-domain correspondence between the sketch and cartoon video frames, and employ a blending module with occlusion estimation to synthesize the middle frame guided by the sketch. After that, the input frames and the synthetic frame equipped with established correspondence are fed into an arbitrary-time frame interpolation pipeline to generate and refine additional inbetween frames. Finally, a module to preserve temporal consistency is employed. Compared to common frame interpolation methods, our approach can address frames with relatively large motion and also has the flexibility to enable users to control the generated video sequences by editing the sketch guidance. By explicitly considering the correspondence between frames and the sketch, we can achieve higher quality results than other image synthesis methods. Our results show that our system generalizes well to different movie frames, achieving better results than existing solutions.

cs.CV↗

arXiv2021

DARTS-: Robustly Stepping out of Performance Collapse Without Indicators

Despite the fast development of differentiable architecture search (DARTS), it suffers from long-standing performance instability, which extremely limits its application. Existing robustifying methods draw clues from the resulting deteriorated behavior instead of finding out its causing factor. Various indicators such as Hessian eigenvalues are proposed as a signal to stop searching before the performance collapses. However, these indicator-based methods tend to easily reject good architectures if the thresholds are inappropriately set, let alone the searching is intrinsically noisy. In this paper, we undertake a more subtle and direct approach to resolve the collapse. We first demonstrate that skip connections have a clear advantage over other candidate operations, where it can easily recover from a disadvantageous state and become dominant. We conjecture that this privilege is causing degenerated performance. Therefore, we propose to factor out this benefit with an auxiliary skip connection, ensuring a fairer competition for all operations. We call this approach DARTS-. Extensive experiments on various datasets verify that it can substantially improve robustness. Our code is available at https://github.com/Meituan-AutoML/DARTS- .

cs.LG↗

arXiv2021

Estimating Optimal Treatment Rules with an Instrumental Variable: A Partial Identification Learning Approach

Individualized treatment rules (ITRs) are considered a promising recipe to deliver better policy interventions. One key ingredient in optimal ITR estimation problems is to estimate the average treatment effect conditional on a subject's covariate information, which is often challenging in observational studies due to the universal concern of unmeasured confounding. Instrumental variables (IVs) are widely-used tools to infer the treatment effect when there is unmeasured confounding between the treatment and outcome. In this work, we propose a general framework of approaching the optimal ITR estimation problem when a valid IV is allowed to only partially identify the treatment effect. We introduce a novel notion of optimality called "IV-optimality". A treatment rule is said to be IV-optimal if it minimizes the maximum risk with respect to the putative IV and the set of IV identification assumptions. We derive a bound on the risk of an IV-optimal rule that illuminates when an IV-optimal rule has favorable generalization performance. We propose a classification-based statistical learning method that estimates such an IV-optimal rule, design computationally-efficient algorithms, and prove theoretical guarantees. We contrast our proposed method to the popular outcome weighted learning (OWL) approach via extensive simulations, and apply our method to study which mothers would benefit from traveling to deliver their premature babies at hospitals with high level neonatal intensive care units.

stat.ME↗

arXiv2021

Robust Dynamical Decoupling for the Manipulation of a Spin Network via a Single Spin

High-fidelity control of quantum systems is crucial for quantum information processing, but is often limited by perturbations from the environment and imperfections in the applied control fields. Here, we investigate the combination of dynamical decoupling (DD) and robust optimal control (ROC) to address this problem. In this combination, ROC is employed to find robust shaped pulses, wherein the directional derivatives of the controlled dynamics with respect to control errors are reduced to a desired order. Then, we incorporate ROC pulses into DD sequences, achieving a remarkable improvement of robustness against multiple error channels. We demonstrate this method in the example of manipulating nuclear spin bath via an electron spin in the NV center system. Simulation results indicate that ROC based DD sequences outperform the state-of-the-art robust DD sequences. Our work has implications for robust quantum control on near-term noisy quantum devices.

quant-ph↗

arXiv2021

Exploring the Galactic Anticenter substructure with LAMOST & Gaia DR2

We characterize the kinematic and chemical properties of 589 Galactic Anticenter Substructure Stars (GASS) with K-/M- giants in Integrals-of-Motion space. These stars likely include members of previously identified substructures such as Monoceros, A13, and the Triangulum-Andromeda cloud (TriAnd). We show that these stars are on nearly circular orbits on both sides of the Galactic plane. We can see velocity($V_{Z}$) gradient along Y-axis especially for the south GASS members. Our GASS members have similar energy and angular momentum distributions to thin disk stars. Their location in [$α$/M] vs. [M/H] space is more metal poor than typical thin disk stars, with [$α$/M] \textbf{lower} than the thick disk. We infer that our GASS members are part of the outer metal-poor disk stars, and the outer-disk extends to 30 kpc. Considering the distance range and $α$-abundance features, GASS could be formed after the thick disk was formed due to the molecular cloud density decreased in the outer disk where the SFR might be less efficient than the inner disk.

astro-ph.GA↗

arXiv2020

HCGrid: A Convolution-based Gridding Framework for RadioAstronomy in Hybrid Computing Environments

Gridding operation, which is to map non-uniform data samples onto a uniformly distributedgrid, is one of the key steps in radio astronomical data reduction process. One of the mainbottlenecks of gridding is the poor computing performance, and a typical solution for suchperformance issue is the implementation of multi-core CPU platforms. Although such amethod could usually achieve good results, in many cases, the performance of gridding is stillrestricted to an extent due to the limitations of CPU, since the main workload of gridding isa combination of a large number of single instruction, multi-data-stream operations, which ismore suitable for GPU, rather than CPU implementations. To meet the challenge of massivedata gridding for the modern large single-dish radio telescopes, e.g., the Five-hundred-meterAperture Spherical radio Telescope (FAST), inspired by existing multi-core CPU griddingalgorithms such as Cygrid, here we present an easy-to-install, high-performance, and open-source convolutional gridding framework, HCGrid,in CPU-GPU heterogeneous platforms. Itoptimises data search by employing multi-threading on CPU, and accelerates the convolutionprocess by utilising massive parallelisation of GPU. In order to make HCGrid a more adaptivesolution, we also propose the strategies of thread organisation and coarsening, as well as optimalparameter settings under various GPU architectures. A thorough analysis of computing timeand performance gain with several GPU parallel optimisation strategies show that it can leadto excellent performance in hybrid computing environments.

astro-ph.IM↗

arXiv2020

Free-Space Optical Communication Using Non-mode-Selective Photonic Lantern Based Coherent Receiver

A free-space optical communication system using non-mode-selective photonic lantern (PL) based coherent receiver is studied. Based on the simulation of photon distribution, the power distribution at the single-mode fiber end of the PL is quantitatively described as a truncated Gaussian distribution over a simplex. The signal-to-noise ratios (SNRs) for the communication system using PL based receiver are analyzed using different combining techniques, including selection combining (SC), equal-gain combining (EGC), and maximal-ratio combining (MRC). The integral solution, series lower bound solution and asymptotic solution are presented for bit-error rate (BER) of PL based receiver, single-mode fiber receiver and multimode fiber receiver over the Gamma-Gamma atmosphere turbulence channels. We demonstrate that the power distribution of the PL has no effect on the SNR and BER performance of the PL based receiver when MRC is used; and it only has limited influence when EGC is used. However, the power distribution of the PL can greatly affect the BER performance when SC is used. Besides, the SNR gains of the PL based receiver using EGC over single-mode fiber receiver and multimode fiber receiver are numerically studied under different imperfect device parameters; and the scope of application of the communication system is further provided.

eess.SP↗

arXiv2020

Oscillatory chiral flows in confined active fluids with obstacles

An active colloidal fluid comprised of self-propelled spinning particles injecting energy and angular momentum at the microscale demonstrates spontaneous collective states that range from flocks to coherent vortices. Despite their seeming simplicity, the emergent far-from-equilibrium behavior of these fluids remains poorly understood, presenting a challenge to the design and control of next-generation active materials. When confined in a ring, such so-called polar active fluids acquire chirality once the spontaneous flow chooses a direction. In a perfect ring, this chirality is indefinitely long-lived. Here, we combine experiments on self-propelled colloidal Quincke rollers and mesoscopic simulations of continuum Toner-Tu equations to explore how such chiral states can be controlled and manipulated by obstacles. For different obstacle geometries three dynamic steady states have been realized: long-lived chiral flow, an apolar state in which the flow breaks up into counter-rotating vortices and an unconventional collective state with flow having an oscillating chirality. The chirality reversal proceeds through the formation of intermittent vortex chains in the vicinity of an obstacle. We demonstrate that the frequency of collective states with oscillating chirality can be tuned by obstacle parameters. We vary obstacle shapes to design chiral states that are independent of initial conditions. Building on our findings, we realize a system with two triangular obstacles that force the active fluid towards a state with a density imbalance of active particles across the ring. Our results demonstrate how spontaneous polar active flows in combination with size and geometry of scatterers can be used to control dynamic patterns of polar active liquids for materials design.

cond-mat.soft↗

arXiv2020

Projecting the optimal control strategy on invasive plants combining effects of herbivores and native plants resistance

Understanding how to limit biological invasion is critical, especially in the context of accelerating anthropogenic ecological changes. Although biological invasion success could be explained by the lack of natural enemies in new regions, recent studies have revealed that resident herbivores often do have a substantial effect on both native and invasive plants. Very few studies have included consideration of native plant resistance while estimating methods of controlling invasion; hence, it is unclear to what extent the interactive effects of controlling approaches and native plants' resistance could slow down or even inhibit biological invasion. We developed a spatial modeling framework, using a paired logistic equation model, with considerations of the dispersal processes, to capture the dynamics change of native and invasive plants under various strategies of control. We found that when biocontrol agents could have a strong effect on invasive plant, that could almost completely limit the invasion, together with a high native plant resistance. However, a high application frequency is needed make an efficient impact, whereas, a low frequency treatment leads to nearly the same outcome as the no treatment case. Lastly, we showed that evenly controlling a larger area with a weaker effect still lead to a better outcome than focusing on small patches with a stronger effect. Overall, this study has some management implications, such as how to determine the optimal allocation strategy.

q-bio.PE↗

arXiv2020

Bi-level Score Matching for Learning Energy-based Latent Variable Models

Score matching (SM) provides a compelling approach to learn energy-based models (EBMs) by avoiding the calculation of partition function. However, it remains largely open to learn energy-based latent variable models (EBLVMs), except some special cases. This paper presents a bi-level score matching (BiSM) method to learn EBLVMs with general structures by reformulating SM as a bi-level optimization problem. The higher level introduces a variational posterior of the latent variables and optimizes a modified SM objective, and the lower level optimizes the variational posterior to fit the true posterior. To solve BiSM efficiently, we develop a stochastic optimization algorithm with gradient unrolling. Theoretically, we analyze the consistency of BiSM and the convergence of the stochastic algorithm. Empirically, we show the promise of BiSM in Gaussian restricted Boltzmann machines and highly nonstructural EBLVMs parameterized by deep convolutional neural networks. BiSM is comparable to the widely adopted contrastive divergence and SM methods when they are applicable; and can learn complex EBLVMs with intractable posteriors to generate natural images.

cs.LG↗

arXiv2020

An approximate factorization method for inverse acoustic scattering with phaseless total-field data

This paper is concerned with the inverse acoustic scattering problem with phaseless total-field data at a fixed frequency. An approximate factorization method is developed to numerically reconstruct both the location and shape of the unknown scatterer from the phaseless total-field data generated by incident plane waves at a fixed frequency and measured on the circle $\partial B_R$ with a sufficiently large radius $R$. The theoretical analysis of our method is based on the asymptotic property in the operator norm from $H^{1/2}(\mathbb{S}^1)$ to $H^{-1/2}(\mathbb{S}^1)$ of the phaseless total-field operator defined in terms of the phaseless total-field data measured on $\partial B_R$ with large enough $R$, where $H^s(\mathbb{S}^1)$ is a Sobolev space on the unit circle $\mathbb{S}^1$ for real number $s$, together with the factorization of a modified far-field operator. The asymptotic property of the phaseless total-field operator is also established in this paper with the theory of oscillatory integrals. The unknown scatterer can be either an impenetrable obstacle of sound-soft, sound-hard or impedance type or an inhomogeneous medium with a compact support, and the proposed inversion algorithm does not need to know the boundary condition of the unknown obstacle in advance. Numerical examples are also carried out to demonstrate the effectiveness of our inversion method. To the best of our knowledge, it is the first attempt to develop a factorization type method for inverse scattering problems with phaseless data.

math.NA↗

arXiv2020

Triple Generative Adversarial Networks

We propose a unified game-theoretical framework to perform classification and conditional image generation given limited supervision. It is formulated as a three-player minimax game consisting of a generator, a classifier and a discriminator, and therefore is referred to as Triple Generative Adversarial Network (Triple-GAN). The generator and the classifier characterize the conditional distributions between images and labels to perform conditional generation and classification, respectively. The discriminator solely focuses on identifying fake image-label pairs. Under a nonparametric assumption, we prove the unique equilibrium of the game is that the distributions characterized by the generator and the classifier converge to the data distribution. As a byproduct of the three-player mechanism, Triple-GAN is flexible to incorporate different semi-supervised classifiers and GAN architectures. We evaluate Triple-GAN in two challenging settings, namely, semi-supervised learning and the extreme low data regime. In both settings, Triple-GAN can achieve excellent classification results and generate meaningful samples in a specific class simultaneously. In particular, using a commonly adopted 13-layer CNN classifier, Triple-GAN outperforms extensive semi-supervised learning methods substantially on more than 10 benchmarks no matter data augmentation is applied or not.

cs.LG↗

arXiv2020

Old Photo Restoration via Deep Latent Space Translation

We propose to restore old photos that suffer from severe degradation through a deep learning approach. Unlike conventional restoration tasks that can be solved through supervised learning, the degradation in real photos is complex and the domain gap between synthetic images and real old photos makes the network fail to generalize. Therefore, we propose a novel triplet domain translation network by leveraging real photos along with massive synthetic image pairs. Specifically, we train two variational autoencoders (VAEs) to respectively transform old photos and clean photos into two latent spaces. And the translation between these two latent spaces is learned with synthetic paired data. This translation generalizes well to real photos because the domain gap is closed in the compact latent space. Besides, to address multiple degradations mixed in one old photo, we design a global branch with apartial nonlocal block targeting to the structured defects, such as scratches and dust spots, and a local branch targeting to the unstructured defects, such as noises and blurriness. Two branches are fused in the latent space, leading to improved capability to restore old photos from multiple defects. Furthermore, we apply another face refinement network to recover fine details of faces in the old photos, thus ultimately generating photos with enhanced perceptual quality. With comprehensive experiments, the proposed pipeline demonstrates superior performance over state-of-the-art methods as well as existing commercial tools in terms of visual quality for old photos restoration.

cs.CV↗