arXiv subjects

Dong Zhang

Publications and source records attributed to Dong Zhang.

At least 199 records · Page 11Linked to original sources

arXiv2020

Resolving Kirchhoff's laws for state-estimator design of Li-ion battery packs connected in parallel

A state-space model for Li-ion battery packs with parallel connected cells is introduced. The key feature of this model is an explicit solution to Kirchhoff's laws for parallel connected packs, which expresses the branch currents directly in terms of the model's states, applied current and cell resistances. This avoids the need to solve these equations numerically. To illustrate the potential of the proposed model for pack-level control and estimation, a state-estimator is introduced for the nonlinear parallel pack model. By exploiting the system structure seen in the solution to Kirchhoff's laws, algebraic conditions for the observer gains are obtained that guarantee convergence of the estimator's error. Error convergence is demonstrated through an argument based upon Aizerman's conjecture. It is hoped that the insight brought by this model formulation will allow the wealth of results developed for series connected packs to be applied to those with parallel connections.

eess.SY↗

arXiv2020

Causal Intervention for Weakly-Supervised Semantic Segmentation

We present a causal inference framework to improve Weakly-Supervised Semantic Segmentation (WSSS). Specifically, we aim to generate better pixel-level pseudo-masks by using only image-level labels -- the most crucial step in WSSS. We attribute the cause of the ambiguous boundaries of pseudo-masks to the confounding context, e.g., the correct image-level classification of "horse" and "person" may be not only due to the recognition of each instance, but also their co-occurrence context, making the model inspection (e.g., CAM) hard to distinguish between the boundaries. Inspired by this, we propose a structural causal model to analyze the causalities among images, contexts, and class labels. Based on it, we develop a new method: Context Adjustment (CONTA), to remove the confounding bias in image-level classification and thus provide better pseudo-masks as ground-truth for the subsequent segmentation model. On PASCAL VOC 2012 and MS-COCO, we show that CONTA boosts various popular WSSS methods to new state-of-the-arts.

cs.CV↗

arXiv2020

Dual-SLAM: A framework for robust single camera navigation

SLAM (Simultaneous Localization And Mapping) seeks to provide a moving agent with real-time self-localization. To achieve real-time speed, SLAM incrementally propagates position estimates. This makes SLAM fast but also makes it vulnerable to local pose estimation failures. As local pose estimation is ill-conditioned, local pose estimation failures happen regularly, making the overall SLAM system brittle. This paper attempts to correct this problem. We note that while local pose estimation is ill-conditioned, pose estimation over longer sequences is well-conditioned. Thus, local pose estimation errors eventually manifest themselves as mapping inconsistencies. When this occurs, we save the current map and activate two new SLAM threads. One processes incoming frames to create a new map and the other, recovery thread, backtracks to link new and old maps together. This creates a Dual-SLAM framework that maintains real-time performance while being robust to local pose estimation failures. Evaluation on benchmark datasets shows Dual-SLAM can reduce failures by a dramatic $88\%$.

cs.RO↗

arXiv2020

Mask Detection and Breath Monitoring from Speech: on Data Augmentation, Feature Representation and Modeling

This paper introduces our approaches for the Mask and Breathing Sub-Challenge in the Interspeech COMPARE Challenge 2020. For the mask detection task, we train deep convolutional neural networks with filter-bank energies, gender-aware features, and speaker-aware features. Support Vector Machines follows as the back-end classifiers for binary prediction on the extracted deep embeddings. Several data augmentation schemes are used to increase the quantity of training data and improve our models' robustness, including speed perturbation, SpecAugment, and random erasing. For the speech breath monitoring task, we investigate different bottleneck features based on the Bi-LSTM structure. Experimental results show that our proposed methods outperform the baselines and achieve 0.746 PCC and 78.8% UAR on the Breathing and Mask evaluation set, respectively.

eess.AS↗

arXiv2020

Direct Quantification for Coronary Artery Stenosis Using Multiview Learning

The quantification of the coronary artery stenosis is of significant clinical importance in coronary artery disease diagnosis and intervention treatment. It aims to quantify the morphological indices of the coronary artery lesions such as minimum lumen diameter, reference vessel diameter, lesion length, and these indices are the reference of the interventional stent placement. In this study, we propose a direct multiview quantitative coronary angiography (DMQCA) model as an automatic clinical tool to quantify the coronary artery stenosis from X-ray coronary angiography images. The proposed DMQCA model consists of a multiview module with two attention mechanisms, a key-frame module, and a regression module, to achieve direct accurate multiple-index estimation. The multi-view module comprehensively learns the Spatio-temporal features of coronary arteries through a three-dimensional convolution. The attention mechanisms of each view focus on the subtle feature of the lesion region and capture the important context information. The key-frame module learns the subtle features of the stenosis through successive dilated residual blocks. The regression module finally generates the indices estimation from multiple features.

eess.IV↗

arXiv2020

Feature Pyramid Transformer

Feature interactions across space and scales underpin modern visual recognition systems because they introduce beneficial visual contexts. Conventionally, spatial contexts are passively hidden in the CNN's increasing receptive fields or actively encoded by non-local convolution. Yet, the non-local spatial interactions are not across scales, and thus they fail to capture the non-local contexts of objects (or parts) residing in different scales. To this end, we propose a fully active feature interaction across both space and scales, called Feature Pyramid Transformer (FPT). It transforms any feature pyramid into another feature pyramid of the same size but with richer contexts, by using three specially designed transformers in self-level, top-down, and bottom-up interaction fashion. FPT serves as a generic visual backbone with fair computational overhead. We conduct extensive experiments in both instance-level (i.e., object detection and instance segmentation) and pixel-level segmentation tasks, using various backbones and head networks, and observe consistent improvement over all the baselines and the state-of-the-art methods.

cs.CV↗

arXiv2020

Geometric photon-drag effect and nonlinear shift current in centrosymmetric crystals

The nonlinear shift current, also known as the bulk photovoltaic current generated by linearly polarized light, has long been known to be absent in crystals with inversion symmetry. Here we argue that a non-zero shift current in centrosymmetric crystals can be activated by a photon-drag effect. Photon-drag shift current proceeds from a `shift current dipole' (a geometric quantity characterizing interband transitions) and manifests a purely transverse response in centrosymmetric crystals. This transverse nature proceeds directly from the shift-vector's pseudovector nature under mirror operation and underscores its intrinsic geometric origin. Photon-drag shift current can greatly enhanced by coupling to polaritons and provides a new and sensitive tool to interrogate the subtle interband coherences of materials with inversion symmetry previously thought to be inaccessible via photocurrent probes.

cond-mat.mes-hall↗

arXiv2020

Reconstructing undersampled photoacoustic microscopy images using deep learning

One primary technical challenge in photoacoustic microscopy (PAM) is the necessary compromise between spatial resolution and imaging speed. In this study, we propose a novel application of deep learning principles to reconstruct undersampled PAM images and transcend the trade-off between spatial resolution and imaging speed. We compared various convolutional neural network (CNN) architectures, and selected a fully dense U-net (FD U-net) model that produced the best results. To mimic various undersampling conditions in practice, we artificially downsampled fully-sampled PAM images of mouse brain vasculature at different ratios. This allowed us to not only definitively establish the ground truth, but also train and test our deep learning model at various imaging conditions. Our results and numerical analysis have collectively demonstrated the robust performance of our model to reconstruct PAM images with as few as 2% of the original pixels, which may effectively shorten the imaging time without substantially sacrificing the image quality.

eess.IV↗

arXiv2020

Khintchine inequality on normed spaces and the application to Banach-Mazur distance

We establish variant Khintchine inequalities on normed spaces of Hanner type and cotype, in which the Rademacher distribution corresponding to classical Khintchine inequality is replaced by general symmetric distributions. The proof involves the $p$-barycenter and Birkhoff's ergodic theorem. More importantly, by employing these Khintchine inequalities, we get some lower bounds for Banach-Mazur distance between $l^p$-ball and a general centrally symmetric convex body.

math.FA↗

arXiv2020

Dusty Cloud Acceleration with Multiband Radiation

We perform two-dimensional and three-dimensional simulations of cold, dense clouds, which are accelerated by radiation pressure on dust relative to a hot, diffuse background gas. We examine the relative effectiveness of acceleration by ultraviolet and infrared radiation fields, both independently and acting simultaneously on the same cloud. We study clouds that are optically thin to infrared emission but with varying ultraviolet optical depths. Consistent with previous work, we find relatively efficient acceleration and long cloud survival times when the infrared band flux dominates over the ultraviolet flux. However, when ultraviolet is dominant or even a modest percentage ($\sim 5-10$\%) of the infrared irradiating flux, it can act to compress the cloud, first crushing it and then disrupting the outer layers. This drives mixing of outer regions of the dusty gas with the hot diffuse background to the point where most dust is not likely to survive or stay coupled to the gas. Hence, the cold cloud is unable to survive for a long enough timescale to experience significant acceleration before disruption even though efficient infrared cooling keeps the majority of the gas close to radiative equilibrium temperature ($T \lesssim 100$K). We discuss implications for observed systems, concluding that radiation pressure driving is most effective when the light from star-forming regions is efficiently reprocessed into the infrared.

astro-ph.GA↗

arXiv2020

Spectrum of signless 1-Laplacian on simplicial complexes

We introduce the signless 1-Laplacians and the dual Cheeger constants on simplicial complexes. The connection of its spectrum to the combinatorial properties like independence number, chromatic number and dual Cheeger constant is investigated. Our estimates can be comparable to Hoffman's bounds in virtue of Laplacian on simplicial complexes. An interesting inequality relating multiplicity of the largest eigenvalue, independence number and chromatic number are provided, which could be regarded as a variant version of Lovasz sandwich theorem. Also, the behavior of the operator under the topological operations of wedge and duplication of motifs is studied. The Courant nodal domain theorem in spectral theory is extended to the case of signless 1-Laplacian on complexes.

math.SP↗

arXiv2020

State of Charge Estimation of Parallel Connected Battery Cells via Descriptor System Theory

This manuscript presents an algorithm for individual Lithium-ion (Li-ion) battery cell state of charge (SOC) estimation when multiple cells are connected in parallel, using only terminal voltage and total current measurements. For battery packs consisting of thousands of cells, it is desirable to estimate individual SOCs by only monitoring the total current in order to reduce sensing cost. Mathematically, series connected cells yield dynamics given by ordinary differential equations under classical full voltage sensing. In contrast, parallel connected cells are evidently more challenging because the dynamics are governed by a nonlinear descriptor system, including differential equations and algebraic equations arising from voltage and current balance across cells. An observer with linear output error injection is formulated, where the individual cell SOCs and local currents are locally observable from the total current and voltage measurements. The asymptotic convergence of differential and algebraic states is established by considering local Lipschitz continuity property of system nonlinearities. Simulation results on LiNiMnCoO$_2$/Graphite (NMC) cells illustrate convergence for SOCs, local currents, and terminal voltage.

eess.SY↗

arXiv2019

A multi-label classification method using a hierarchical and transparent representation for paper-reviewer recommendation

Paper-reviewer recommendation task is of significant academic importance for conference chairs and journal editors. How to effectively and accurately recommend reviewers for the submitted papers is a meaningful and still tough task. In this paper, we propose a Multi-Label Classification method using a hierarchical and transparent Representation named Hiepar-MLC. Further, we propose a simple multi-label-based reviewer assignment MLBRA strategy to select the appropriate reviewers. It is interesting that we also explore the paper-reviewer recommendation in the coarse-grained granularity.

cs.IR↗

arXiv2019

Electrical Control of Large Rashba Effect in Oxide Heterostructures

Large Rashba effect efficiently tuned by an external electric field is highly desired for spintronic devices. Using first-principles calculations, we demonstrate that large Rashba splitting is locked at conduction band minimum in ferroelectric Bi(Sc/Y/La/Al/Ga/In)O3/PbTiO3 heterostructures where the position of Fermi level is precisely controlled via its stoichiometry. Fully reversible Rashba spin texture and drastic change of Rashba splitting strength with ferroelectric polarization switching are realized in the symmetric and asymmetric heterostructures, respectively. By artificially tuning the local ferroelectric displacement and the orbital hybridization, the synergetic effect of local potential gradient and orbital overlap on the dramatic change of splitting strength is confirmed. These results improve the feasibility of utilizing Rashba spin-orbit coupling in spintronic devices.

cond-mat.mtrl-sci↗

arXiv2019

Lovász extension and graph cut

A set-pair Lovász extension is established to construct equivalent continuous optimization problems for graph $k$-cut problems.

math.CO↗

arXiv2019

Few-shot machine learning in the three-dimensional Ising model

We investigate theoretically the phase transition in three dimensional cubic Ising model utilizing state-of-the-art machine learning algorithms. Supervised machine learning models show high accuracies (~99\%) in phase classification and very small relative errors ($< 10^{-4}$) of the energies in different spin configurations. Unsupervised machine learning models are introduced to study the spin configuration reconstructions and reductions, and the phases of reconstructed spin configurations can be accurately classified by a linear logistic algorithm. Based on the comparison between various machine learning models, we develop a few-shot strategy to predict phase transitions in larger lattices from trained sample in smaller lattices. The few-shot machine learning strategy for three dimensional(3D) Ising model enable us to study 3D ising model efficiently and provides a new integrated and highly accurate approach to other spin models.

cond-mat.dis-nn↗

arXiv2018

Border Effect of Complex Network: An analysis on the cooperation network of movie stars across different regions

People always carry out social activities within certain geographic space. The geographic boundary likes national boundaries and provincial boundaries functions as dividing policies, distributing administrative resources, defining the scope of application of laws and regulations or even the boundary of cultural region, producing fundamental influence on lifestyle. Network analysis is a powerful tool to abstract complex system. In this article, we used network method to understand changes of cooperation network in movie industry and effect of geographical boundary.

cs.CY↗

arXiv2018

A Review of the Theory of Galactic Winds Driven by Stellar Feedback

Galactic winds from star-forming galaxies are crucial to the process of galaxy formation and evolution, regulating star formation, shaping the stellar mass function and the mass-metallicity relation, and enriching the intergalactic medium with metals. Galactic winds associated with stellar feedback may be driven by overlapping supernova explosions, radiation pressure of starlight on dust grains, and cosmic rays. Galactic winds are multiphase, the growing observations of emission and absorption of cold molecular, cool atomic, ionized warm and hot outflowing gas in a large number of galaxies have not been completely understood. In this review article, I summarize the possible mechanisms associated with stars to launch galactic winds, and review the multidimensional hydrodynamic, radiation hydrodynamic and magnetohydrodynamic simulations of winds based on various algorithms. I also briefly discuss the theoretical challenges and possible future research directions.

astro-ph.GA↗