December 2020 arXiv papers — page 5

Shaobo Li, Xiaoguang Li, Lifeng Shang, Xin Jiang

Collecting supporting evidence from large corpora of text (e.g., Wikipedia) is of great challenge for open-domain Question Answering (QA). Especially, for multi-hop open-domain QA, scattered evidence pieces are required to be gathered together to support the answer extraction. In this paper, we propose a new retrieval target, hop, to collect the hidden reaso

Ofir T. Erlich, David H. Lorenz

A software architect turns system requirements into a suitable software architecture through an architecture optimization process. However, how should the architect decide which quality improvement to prioritize, e.g., security or reliability? In software product line, should a small improvement in multiple products be preferred over a large improvement in a

Luc Trouche

The Documentational Approach To Didactics (DAD) aims to study teachers' professional development through their interactions with their resources for/from teaching. It has been introduced in the French community of didactics of mathematics in 2007, then extended at an international level. It has been introduced as an entry in the Encyclopedia of Mathematics E

Zhi-Qin Zhan, Huazhu Fu, Yan-Yao Yang, Jingjing Chen

Automatic colorectal polyp detection in colonoscopy video is a fundamental task, which has received a lot of attention. Manually annotating polyp region in a large scale video dataset is time-consuming and expensive, which limits the development of deep learning techniques. A compromise is to train the target model by using labeled images and infer on colono

Elefterios Soultanis, Stefan Wenger

We introduce and study a notion of relative 1-homotopy type for Sobolev maps from a surface to a metric space spanning a given collection of Jordan curves. We use this to establish the existence and local H\"older regularity of area minimizing surfaces in a given relative 1-homotopy class in proper geodesic metric spaces admitting a local quadratic isoperime

Andrey Levin, Mikhail Olshanetsky, Andrei Zotov

We introduce a notion of quasi-antisymmetric Higgs $G$-bundles over curves with marked points. They are endowed with additional structures, which replace the parabolic structures at marked points in the parabolic Higgs bundles. The latter means that the coadjoint orbits are attached to the marked points. The moduli spaces of parabolic Higgs bundles are the p

Sébastien Biebler

A blender for a surface endomorphism is a hyperbolic basic set for which the union of the local unstable manifolds contains robustly an open set. Introduced by Bonatti and D{\'i}az in the 90s, blenders turned out to have many powerful applications to differentiable dynamics. In particular, a generalization in terms of jets, called parablenders, allowed Berge

José A. Carrillo, Lin Chen, Qi Wang

We propose and analyze an optimal mass transport method for a random genetic drift problem driven by a Moran process under weak-selection. The continuum limit, formulated as a reaction-advection-diffusion equation known as the Kimura equation, inherits degenerate diffusion from the discrete stochastic process that conveys to the blow-up into Dirac-delta sing

Mikhail Chebunin, Artyom Kovalevskii

We prove theorems about the Gaussian asymptotics of an empirical bridge built from linear model regressors with multiple regressor ordering. We study the testing of the hypothesis of a linear model for the components of a random vector: one of the components is a linear combination of the others up to an error that does not depend on the other components of

Weizhen Qi, Yeyun Gong, Jian Jiao, Yu Yan

In this paper, we propose BANG, a new pretraining model to Bridge the gap between Autoregressive (AR) and Non-autoregressive (NAR) Generation. AR and NAR generation can be uniformly regarded as to what extent previous tokens can be attended, and BANG bridges AR and NAR generation by designing a novel model structure for large-scale pretraining. The pretraine

Xinying Song, Alex Salcianu, Yang Song, Dave Dopson

Tokenization is a fundamental preprocessing step for almost all NLP tasks. In this paper, we propose efficient algorithms for the WordPiece tokenization used in BERT, from single-word tokenization to general text (e.g., sentence) tokenization. When tokenizing a single word, WordPiece uses a longest-match-first strategy, known as maximum matching. The best kn

Ying-Ming Xie, Xue-Jian Gao, Tai-Kai Ng, K. T. Law

In this work, we demonstrate that making a cut (a narrow vacuum regime) in the bulk of a quantum anomalous Hall insulator (QAHI) creates a topologically protected single helical channel with counter-propagating electron modes, and inducing superconductivity on the helical channel through proximity effect will create Majorana zero energy modes (MZMs) at the e

Hongjian Wang, Qi Li, Lanbo Zhang, Yue Lu

Historical features are important in ads click-through rate (CTR) prediction, because they account for past engagements between users and ads. In this paper, we study how to efficiently construct historical features through counting features. The key challenge of such problem lies in how to automatically identify counting keys. We propose a tree-based method

A. Pikalov

In this paper we consider holographic model of exciton condensation in double monolayer Dirac semimetal. Excitons is a bound states of an electron and a hole. Being Bose particles, excitons can form a Bose-Einstein condensate. We study formation of two types of condensates. In first case both the electron and the hole forming the exciton are in the same laye

Wissam Antoun, Fady Baly, Hazem Hajj

Recently, pre-trained transformer-based architectures have proven to be very efficient at language modeling and understanding, given that they are trained on a large enough corpus. Applications in language generation for Arabic are still lagging in comparison to other NLP advances primarily due to the lack of advanced Arabic language generation models. In th

Milad Malekzadeh, Ioannis Papamichail, Markos Papageorgiou

Lane-free vehicle movement has been recently proposed for connected automated vehicles (CAV) due to various potential advantages. One such advantage stems from the fact that incremental changes of the road width in lane-free traffic lead to corresponding incremental changes of the traffic flow capacity. Based on this property, the concept of internal boundar

Serhad Sarica, Jianxi Luo

Engineers often need to discover and learn designs from unfamiliar domains for inspiration or other particular uses. However, the complexity of the technical design descriptions and the unfamiliarity to the domain make it hard for engineers to comprehend the function, behavior, and structure of a design. To help engineers quickly understand a complex technic

Sergey Sadov

We systematically explore a class of constrained optimization problems with linear objective function and constraints that are linear combinations of logarithms of the optimization variables. Such problems can be viewed as a generalization of the inequality between the arithmetic and geometric means. The existence and uniqueness of the minimizer is proved un

Wissam Antoun, Fady Baly, Hazem Hajj

Advances in English language representation enabled a more sample-efficient pre-training task by Efficiently Learning an Encoder that Classifies Token Replacements Accurately (ELECTRA). Which, instead of training a model to recover masked tokens, it trains a discriminator model to distinguish true input tokens from corrupted tokens that were replaced by a ge

Zhixing Tan, Shuo Wang, Zonghan Yang, Gang Chen

Machine translation (MT) is an important sub-field of natural language processing that aims to translate natural languages using computers. In recent years, end-to-end neural machine translation (NMT) has achieved great success and has become the new mainstream method in practical MT systems. In this article, we first provide a broad review of the methods fo

Jaeseop Ahn, Jimyeong Kim, Ihyeok Seo

In this paper, we study the persistence of spatial analyticity for the solutions to the Klein-Gordon-Schr\"{o}dinger system, which describes a physical system of a nucleon field interacting with a neutral meson field, with analytic initial data. Unlike the case of a single nonlinear dispersive equation, not much is known about nonlinear dispersive systems as

Yun-Feng Liang, Xing-Fu Zhang, Ji-Gui Cheng, Hou-Dun Zeng

Axion-like particles (ALPs) provide a feasible explanation for the observed lower TeV opacity of the Universe. If the anomaly TeV transparency is caused by ALPs, then the fluxes of distant extragalactic sources will be enhanced at photon energies beyond TeV, resulting in an enhancement of the observed extragalactic gamma-ray background (EGB) spectrum. In thi

Asghar Ullah

In this paper, the dynamics of quantum Fisher information of a qubit interacting with a squeezed thermal environment are studied. The optimal initial state of the qubit, the temperature of the environment, and the interaction time, which maximize quantum Fisher information are obtained. Based on the ohmicity of the environment, we compare the dynamics of qua

Han Shen, Kaiqing Zhang, Mingyi Hong, Tianyi Chen

Asynchronous and parallel implementation of standard reinforcement learning (RL) algorithms is a key enabler of the tremendous success of modern RL. Among many asynchronous RL algorithms, arguably the most popular and effective one is the asynchronous advantage actor-critic (A3C) algorithm. Although A3C is becoming the workhorse of RL, its theoretical proper

Yang Han

We give two sufficient and necessary conditions for a Hochschild extension of a finite dimensional algebra by its dual bimodule and a Hochschild 2-cocycle to be a symmetric algebra.

Atila Abdulkadiroglu, Joshua D. Angrist, Yusuke Narita, Parag Pathak

Many schools in large urban districts have more applicants than seats. Centralized school assignment algorithms ration seats at over-subscribed schools using randomly assigned lottery numbers, non-lottery tie-breakers like test scores, or both. The New York City public high school match illustrates the latter, using test scores and other criteria to rank app

Yuanlin Li, Qinghai Zhong

A ring $R$ is said to be clean if each element of $R$ can be written as the sum of a unit and an idempotent. $R$ is said to be weakly clean if each element of $R$ is either a sum or a difference of a unit and an idempotent, and $R$ is said to be feebly clean if every element $r$ can be written as $r=u+e_1-e_2$, where $u$ is a unit and $e_1,e_2$ are orthogona

Chung-Chuan Chen, Lin-Ling Huang

In this note, we obtain some identities for the generalized Fibonacci polynomial by using the Q(x) matrix. These identities including the Cassini identity and Honsberger formula can be applied to some polynomial sequences, such as Fibonacci polynomials, Lucas polynomials, Pell polynomials, Pell-Lucas polynomials, Fermat polynomials, Fermat-Lucas polynomials,

Jacques Benatar, Alon Nishry, Brad Rodgers

For $X(n)$ a Rademacher or Steinhaus random multiplicative function, we consider the random polynomials $$ P_N(\theta) = \frac1{\sqrt{N}} \sum_{n\leq N} X(n) e(n\theta), $$ and show that the $2k$-th moments on the unit circle $$ \int_0^1 \big| P_N(\theta) \big|^{2k}\, d\theta $$ tend to Gaussian moments in the sense of mean-square convergence, uniformly for

Lijun Yuan, Ya Yan Lu

Bound states in the continuum (BICs) are trapped or guided modes with frequencies in radiation continua. They are associated with high-quality-factor resonances that give rise to strong local field enhancement and rapid variations in scattering spectra, and have found many valuable applications. A guided mode of an optical waveguide can also be a BIC, if the

Anthony D. Blaom, Sebastian J. Vollmer

A graph-based protocol called `learning networks' which combine assorted machine learning models into meta-models is described. Learning networks are shown to overcome several limitations of model composition as implemented in the dominant machine learning platforms. After illustrating the protocol in simple examples, a concise syntax for specifying a learni

Andrea Madotto, Zhaojiang Lin, Zhenpeng Zhou, Seungwhan Moon

Continual learning in task-oriented dialogue systems can allow us to add new domains and functionalities through time without incurring the high cost of a whole system retraining. In this paper, we propose a continual learning benchmark for task-oriented dialogue systems with 37 domains to be learned continuously in four settings, such as intent recognition,

Lianli Gao, Qilong Zhang, Jingkuan Song, Heng Tao Shen

Although great progress has been made on adversarial attacks for deep neural networks (DNNs), their transferability is still unsatisfactory, especially for targeted attacks. There are two problems behind that have been long overlooked: 1) the conventional setting of $T$ iterations with the step size of $\epsilon/T$ to comply with the $\epsilon$-constraint. I

Itai Benjamini, Mikolaj Fraczyk, Gabor Kun

We conjecture that finite graphs with positive Cheeger constant admit a spanning subgraph with positive Cheeger constant and girth proportional to the diameter. We prove this conjecture for regular expander graphs with large expansion. Our proof relies on the Local Lemma.

Zhaoyu Fei, C. P. Sun

In this paper, we systematically study the work statistics for quantum phase transition. For a quantum system approached by an anisotropic conformal field theory near the critical point, the driving protocols is divided into three different regimes for different quench rates, which reflects the competition between the frozen time and the quench time scale. I

Dao-Sheng Tang, Bing-Yang Cao

Tuning thermal transport in semiconductor nanostructures is of great significance for thermal management in information and power electronics. With excellent transport properties, such as ballistic transport, immunity to point defects and disorders, and forbidden backscattering, topological phonon surface states show remarkable potential in addressing this i

Alessio Figalli, Sunghan Kim, Henrik Shahgholian

We consider vector-valued solutions to a linear transmission problem, and we prove that Lipschitz-regularity on one phase is transmitted to the next phase. More exactly, given a solution $u:B_1\subset \mathbb{R}^n \to \mathbb{R}^m$ to the elliptic system \begin{equation*} \mbox{div} ((A + (B-A)\chi_D )\nabla u) = 0 \quad \text{in }B_1, \end{equation*} where

Chien-Ming Lin, Yu-Ming Hsu, Yen-Huan Li

Quantum state tomography (QST), the task of estimating an unknown quantum state given measurement outcomes, is essential to building reliable quantum computing devices. Whereas computing the maximum-likelihood (ML) estimate corresponds to solving a finite-sum convex optimization problem, the objective function is not smooth nor Lipschitz, so most existing co

Kun Wei, Cheng Deng, Xu Yang, Maosen Li

Modern deep learning methods have achieved great success in machine learning and computer vision fields by learning a set of pre-defined datasets. Howerver, these methods perform unsatisfactorily when applied into real-world situations. The reason of this phenomenon is that learning new tasks leads the trained model quickly forget the knowledge of old tasks,

Yiling Yang, Qiaoyuan Cheng, Engui Fan

In this paper, we apply $\overline\partial$ steepest descent method to study the Cauchy problem for the derivative nonlinear Schr\"odinger equation with nonzero boundary conditions \begin{align} &iq_{t}+q_{xx}+i\sigma(|q|^2q)_{x}=0,\\ & (x,0) = q_0(x), \quad\lim_{x\to\pm\infty} q_0(x) = q_\pm,\end{align} where $|q_\pm|=1$. Based on the spectral analysis of t

Ahmad Rashid, Vasileios Lioutas, Abbas Ghaddar, Mehdi Rezagholizadeh

Knowledge Distillation (KD) is a common knowledge transfer algorithm used for model compression across a variety of deep learning based natural language processing (NLP) solutions. In its regular manifestations, KD requires access to the teacher's training data for knowledge transfer to the student network. However, privacy concerns, data regulations and pro

Furu Zhang, Yang Gao, Wei Zhang

We systematically investigate the properties of bulk, surface and edge plasmons in Weyl semimetals in presence of a magnetic field. It is found that unidirectional plasmons with different properties exist on different surfaces, which is in consistent with the nontrivial topology of the three-dimensional (3D) bulk plasmons. These novel plasmons possess moment

Boris Skoric

We introduce a variant of quantum signatures in which nonbinary symbols are signed instead of bits. The public keys are fingerprinting states, just as in the scheme of Gottesman and Chuang, but we allow for multiple ways to reveal the private key partially. The effect of this modification is a reduction of the number of qubits expended per message bit. We gi

Rahman Salim Zengin, Volkan Sezer

Voronoi tessellations are used to partition the Euclidean space into polyhedral regions, which are called Voronoi cells. Labeling the Voronoi cells with the class information, we can map any classification problem into a Voronoi tessellation. In this way, the classification problem changes into a query of just finding the enclosing Voronoi cell. In order to

Anton Kapustin, Nikita Sopenko, Bowen Yang

We study invertible states of 1d bosonic quantum lattice systems. We show that every invertible 1d state is in a trivial phase: after tensoring with some unentangled ancillas it can be disentangled by a fuzzy analog of a finite-depth quantum circuit. If an invertible state has symmetries, it may be impossible to disentangle it in a way that preserves the sym

Vladimir Karas, Ondrej Kopacek

We study the effects of large scale magnetic fields on the dynamics of charged particles near a rotating black hole. We consider a scenario in which the initially neutral particles on geodesic orbits in the equatorial plane become ionized, and hence they are destabilized by the charging process. Fraction of charged particles are then accelerated out of the e

Yeting Li, Shuaimin Li, Zhiwu Xu, Jialun Cao

Since regular expressions (abbrev. regexes) are difficult to understand and compose, automatically generating regexes has been an important research problem. This paper introduces TransRegex, for automatically constructing regexes from both natural language descriptions and examples. To the best of our knowledge, TransRegex is the first to treat the NLP-and-

Hannah Lu, Dinara Ermakova, Haruko Murakami Wainwright, Liange Zheng

Machine learning techniques are powerful tools for construction of emulators for complex systems. We explore different machine learning methods and conceptual methodologies, ranging from functional approximations to dynamical approximations, to build such emulators for coupled thermal, hydrological, mechanical and chemical processes that occur near an engine

Sung-Gook Choi, Deok-Sun Lee

The distributions of trade values and relationships among countries and product categories reflect how countries select their trade partners and design export portfolios. Here we consider the exporter-importer network and the exporter-product network with directed links weighted by the logarithm of the corresponding export values each year from 1962 to 2018,

Seunghoon Lee, Chanho Park, Song-Nam Hong, Yonina C. Eldar

Federated learning is a privacy-preserving and distributed training method using heterogeneous data sets stored at local devices. Federated learning over wireless networks requires aggregating locally computed gradients at a server where the mobile devices send statistically distinct gradient information over heterogenous communication links. This paper prop

Mahsa Ghasemi, Evan Scope Crafts, Bo Zhao, Ufuk Topcu

In planning problems, it is often challenging to fully model the desired specifications. In particular, in human-robot interaction, such difficulty may arise due to human's preferences that are either private or complex to model. Consequently, the resulting objective function can only partially capture the specifications and optimizing that may lead to poor

Kiran Ramnath, Mark Hasegawa-Johnson

Fact-based Visual Question Answering (FVQA), a challenging variant of VQA, requires a QA-system to include facts from a diverse knowledge graph (KG) in its reasoning process to produce an answer. Large KGs, especially common-sense KGs, are known to be incomplete, i.e., not all non-existent facts are always incorrect. Therefore, being able to reason over inco

Horia Mania, Suvrit Sra

Recent studies of generalization in deep learning have observed a puzzling trend: accuracies of models on one data distribution are approximately linear functions of the accuracies on another distribution. We explain this trend under an intuitive assumption on model similarity, which was verified empirically in prior work. More precisely, we assume the proba

Kushal Lakhotia, Bhargavi Paranjape, Asish Ghoshal, Wen-tau Yih

Natural language (NL) explanations of model predictions are gaining popularity as a means to understand and verify decisions made by large black-box pre-trained models, for NLP tasks such as Question Answering (QA) and Fact Verification. Recently, pre-trained sequence to sequence (seq2seq) models have proven to be very effective in jointly making predictions

Ari Arapostathis, Anup Biswas, Somnath Pradhan

We consider general linear non-degenerate weakly-coupled cooperative elliptic systems and study certain monotonicity properties of the generalized principal eigenvalue in $\mathbb{R}^d$ with respect to the potential. It is shown that monotonicity on the right is equivalent to the recurrence property of the twisted operator which is, in turn, equivalent to th

Rob Brekelmans, Frank Nielsen, Alireza Makhzani, Aram Galstyan

The exponential family is well known in machine learning and statistical physics as the maximum entropy distribution subject to a set of observed constraints, while the geometric mixture path is common in MCMC methods such as annealed importance sampling. Linking these two ideas, recent work has interpreted the geometric mixture path as an exponential family

M M Akbar, R Solanki

We study the Einstein equations of the static spherically symmetric anisotropic fluid system in curvature coordinates to find algorithms that generate all solutions and all solutions that are regular at the center. All possible combinations of input functions from the set of four functions that characterize the anisotropic system are considered and all equiv

Blake Mellor, Riley Smith

The fundamental quandle is a powerful invariant of knots and links, but it is difficult to describe in detail. It is often useful to look at quotients of the quandle, especially finite quotients. One natural quotient introduced by Joyce is the $n$-quandle. Hoste and Shanahan gave a complete list of the knots and links which have finite $n$-quandles for some

Atsushi Nitanda, Denny Wu, Taiji Suzuki

We propose the particle dual averaging (PDA) method, which generalizes the dual averaging method in convex optimization to the optimization over probability distributions with quantitative runtime guarantee. The algorithm consists of an inner loop and outer loop: the inner loop utilizes the Langevin algorithm to approximately solve for a stationary distribut

Shashwat Singh, Ankul Prajapati, Kamlesh N Pathak

In a quest towards an intelligent decision-making machine, the ability to make plausible predictions is the central pillar of its intelligence. A predicting algorithm's central idea is to understand the governing physical rules and make plausible and apt predictions based on the same governing laws. Extending the study towards the astrophysical phenomenon pu

Arijit Dutta, Jaewan Kim, Jinhyoung Lee

We introduce two types of statistical quasi-separation between local observables to construct two-party Bell-type inequalities for an arbitrary dimensional systems and arbitrary number of measurement settings per site. Note that, the main difference between statistical quasi-separations and the usual statistical separations is that the former are not symmetr

H. H. Jen, J. -S. You

An atom-waveguide system, which presents one of the quantum interfaces that enable strong couplings between light and atoms, can support tightly-confined guided modes of light. In this distinctive quantum interface, we theoretically investigate the crossover from a delocalized to localized atomic excitation under long-range dipole-dipole interactions and lat

Lianjin Wu

BESIII experiment has collected about 20 fb$^{-1}$ luminosity data between $\sqrt{s}=3.8$ and 4.7 GeV via $e^{+}e^{-}$ collision. In this talk, we present the recent $XYZ$ results at BESIII, including mass and width measurements of $Y(4220)$, search for $Y$ states with rare decay modes, isospin parity determination of $Z_{c}^{0}(3900)$ as well as the mass an

Yoav Alon, Huiyu Zhou

Recent technological progress in the development of Unmanned Aerial Vehicles (UAVs) together with decreasing acquisition costs make the application of drone fleets attractive for a wide variety of tasks. In agriculture, disaster management, search and rescue operations, commercial and military applications, the advantage of applying a fleet of drones origina

Eero Siivola, Javier Gonzalez, Andrei Paleyes, Aki Vehtari

The increasing availability of structured but high dimensional data has opened new opportunities for optimization. One emerging and promising avenue is the exploration of unsupervised methods for projecting structured high dimensional data into low dimensional continuous representations, simplifying the optimization problem and enabling the application of tr

Senthil Purushwalkam, Sebastian Vicenc Amengual Gari, Vamsi Krishna Ithapu, Carl Schissler

Given only a few glimpses of an environment, how much can we infer about its entire floorplan? Existing methods can map only what is visible or immediately apparent from context, and thus require substantial movements through a space to fully map it. We explore how both audio and visual sensing together can provide rapid floorplan reconstruction from limited

Tianyi Chen, Ziye Guo, Yuejiao Sun, Wotao Yin

Stochastic gradient descent (SGD) has taken the stage as the primary workhorse for large-scale machine learning. It is often used with its adaptive variants such as AdaGrad, Adam, and AMSGrad. This paper proposes an adaptive stochastic gradient descent method for distributed machine learning, which can be viewed as the communication-adaptive counterpart of t

Minyi Huang, Xuwei Yang

This paper studies asymptotic solvability of a linear quadratic (LQ) mean field social optimization problem with controlled diffusions and indefinite state and control weights. Starting with an $N$-agent model, we employ a rescaling approach to derive a low-dimensional Riccati ordinary differential equation (ODE) system, which characterizes a necessary and s

Thomas Y. Hou, Zhenzhen Li, Ziyun Zhang

In this paper, we propose a new global analysis framework for a class of low-rank matrix recovery problems on the Riemannian manifold. We analyze the global behavior for the Riemannian optimization with random initialization. We use the Riemannian gradient descent algorithm to minimize a least squares loss function, and study the asymptotic behavior as well

Zhuofeng Wu, Sinong Wang, Jiatao Gu, Madian Khabsa

Pre-trained language models have proven their unique powers in capturing implicit language features. However, most pre-training approaches focus on the word-level training objective, while sentence-level objectives are rarely studied. In this paper, we propose Contrastive LEArning for sentence Representation (CLEAR), which employs multiple sentence-level aug

Hirohisa Watanabe, Hiroki Matsutani

ODENet is a deep neural network architecture in which a stacking structure of ResNet is implemented with an ordinary differential equation (ODE) solver. It can reduce the number of parameters and strike a balance between accuracy and performance by selecting a proper solver. It is also possible to improve the accuracy while keeping the same number of paramet

Papri Dey, Hema Srinivasan

Principal matrices of a numerical semigroup of embedding dimension n are special types of $n \times n$ matrices over integers of rank $\leq n - 1$. We show that such matrices and even the pseudo principal matrices of size n must have rank $\geq \frac{n}{2}$ regardless of the embedding dimension. We give structure theorems for pseudo principal matrices for wh

Mohammad Akbari, Jie Liang, Jingning Han, Chengjie Tu

Recently deep learning-based image compression has shown the potential to outperform traditional codecs. However, most existing methods train multiple networks for multiple bit rates, which increase the implementation complexity. In this paper, we propose a new variable-rate image compression framework, which employs generalized octave convolutions (GoConv)

Dan Lin, Jiajing Wu, Qi Yuan, Zibin Zheng

As the largest public blockchain-based platform supporting smart contracts, Ethereum has accumulated a large number of user transaction records since its debut in 2014. Analysis of Ethereum transaction records, however, is still relatively unexplored till now. Modeling the transaction records as a static simple graph, existing methods are unable to accuratel

Sipu Ruan, Gregory S. Chirikjian

This article derives closed-form parametric formulas for the Minkowski sums of convex bodies in d-dimensional Euclidean space with boundaries that are smooth and have all positive sectional curvatures at every point. Under these conditions, there is a unique relationship between the position of each boundary point and the surface normal. The main results are

Peize Sun, Jinkun Cao, Yi Jiang, Rufeng Zhang

In this work, we propose TransTrack, a simple but efficient scheme to solve the multiple object tracking problems. TransTrack leverages the transformer architecture, which is an attention-based query-key mechanism. It applies object features from the previous frame as a query of the current frame and introduces a set of learned object queries to enable detec

Some Sankar Bhattacharya, Ananda G. Maity, Tamal Guha, Giulio Chiribella

We introduce the task of random-receiver quantum communication, in which a sender transmits a quantum message to a receiver chosen from a list of n spatially separated parties. The choice of receiver is unknown to the sender, but is known by the n parties, who coordinate their actions by exchanging classical messages. In normal conditions, random-receiver qu

Vincent Roulet, Zaid Harchaoui

The notion of a Moreau envelope is central to the analysis of first-order optimization algorithms for machine learning. Yet, it has not been developed and extended to be applied to a deep network and, more broadly, to a machine learning system with a differentiable programming implementation. We define a compositional calculus adapted to Moreau envelopes and

Yinan Zheng, Qian Xiao

In this paper we introduce the definition of topological $r$-pressure of free semigroup actions on compact metric space and provide some properties of it. Through skew-product transformation into a medium, we can obtain the following two main results. 1. We extend the result that the topological pressure is the limit of topological $r$-pressure in\cite{C} to

Chin-Yu Hsiao, Wei-Chuan Shen

In this paper, we calculate the second coefficient of the asymptotic expansion of Boutet de Monvel--Sj\"ostrand.

Alham Fikri Aji, Kenneth Heafield

This paper explores augmenting monolingual data for knowledge distillation in neural machine translation. Source language monolingual text can be incorporated as a forward translation. Interestingly, we find the best way to incorporate target language monolingual text is to translate it to the source language and round-trip translate it back to the target la

Wei-Ning Hsu, David Harwath, Christopher Song, James Glass

In this paper we present the first model for directly synthesizing fluent, natural-sounding spoken audio captions for images that does not require natural language text as an intermediate representation or source of supervision. Instead, we connect the image captioning module and the speech synthesis module with a set of discrete, sub-word speech units that

Cris Negron, Julia Pevtsova

We introduce an infinite variant of hypersurface support for finite-dimensional, noncommutative complete intersections. By a noncommutative complete intersection we mean an algebra R which admits a smooth deformation $Q\to R$ by a Noetherian algebra $Q$ which is of finite global dimension. We show that hypersurface support defines a support theory for the bi

Jem N. Corcoran, Caleb Miller

Order statistics arising from $m$ independent but not identically distributed random variables are typically constructed by arranging some $X_{1}, X_{2}, \ldots, X_{m}$, with $X_{i}$ having distribution function $F_{i}(x)$, in increasing order denoted as $X_{(1)} \leq X_{(2)} \leq \ldots \leq X_{(m)}$. In this case, $X_{(i)}$ is not necessarily associated wi

A. W. Bray, C. Simenel

Background: Idealised systems are commonly used in nuclear physics and condensed matter. For instance, the construction of nuclear energy density functionals involves properties of infinite matter, while neutron drops are used to test nuclear interactions and approximations to the nuclear many-body problem. In condensed matter, quantum rings are also used to

Milad Soleimani, Mladen Kezunovic

High penetration of plug-in electric vehicles (PEVs) can potentially put the utility assets such as transformer under overload stress causing decrease in their lifetime. The decrease in PV and battery energy storage system (BESS) prices has made them viable solutions to mitigate this situation. In this paper, the economic aspect of their optimal coordination

Mark Lam, Sambit B. Pal, Thibault Vogt, Martin Kiffner

We optically excite $^{85}$Rb atoms in a heated vapor cell to a low-lying Rydberg state 10D$_{5/2}$ and observe directional terahertz (THz) beams at 3.3 THz and 7.8 THz. These THz fields are generated by amplified spontaneous emission from the 10D$_{5/2}$ state to the 11P$_{3/2}$ and 8F$_{7/2}$ states, respectively. In addition, we observe ultraviolet (UV) l

Vladimir Pletser

We search for triangular numbers that are multiples of other triangular numbers. It is found that for any positive non-square integer multiplier, there is an infinity of multiples of triangular numbers that are triangular numbers and recurrent relations are deduced theoretically. If the multiplier is a squared integer, there is either one or no solution, dep

A. Ibragimov, A. Peace

We develop and analyze a spatial temporal model of light driven ecotoxicological processes, motivated by an aquatic predator-prey system of algae and \textsl{Daphnia} subject to a contaminant. Population dynamics are driven by light, which is periodic in time and varies with spatial depth. The existence and uniqueness of spatial and temporal dependent period

Selvi Kara, Kuei-Nuan Lin, Gabriel Sosa

We prove that the multi-Rees algebra $\mathcal{R}(I_1 \oplus \cdots \oplus I_r)$ of a collection of strongly stable ideals $I_1, \ldots, I_r$ is of fiber type. In particular, we provide a Gr\"obner basis for its defining ideal as a union of a Gr\"obner basis for its special fiber and binomial syzygies. We also study the Koszulness of $\mathcal{R}(I_1 \oplus

Yang Lei, Richard L. J. Qiu, Tonghe Wang, Walter J. Curran

This chapter reviews recent developments of generative adversarial networks (GAN)-based methods for medical and biomedical image synthesis tasks. These methods are classified into conditional GAN and Cycle-GAN according to the network architecture designs. For each category, a literature survey is given, which covers discussions of the network architecture d

Hao Yuan, Haiyang Yu, Shurui Gui, Shuiwang Ji

Deep learning methods are achieving ever-increasing performance on many artificial intelligence tasks. A major limitation of deep models is that they are not amenable to interpretability. This limitation can be circumvented by developing post hoc techniques to explain the predictions, giving rise to the area of explainability. Recently, explainability of dee

Rajesh Ranjan, J. -Ch. Robinet, Datta Gaitonde

Wakes of upswept afterbodies are often characterized by a counter-rotating streamwise vortex pair. The unsteady dynamics of these vortices are examined with a spatio-temporally resolved Large-Eddy Simulation dataset on a representative configuration consisting of a cylinder with an upswept basal surface. Emphasis is placed on understanding the meandering mot

Jiasi Shen, Martin Rinard, Nikos Vasilakis

We present KumQuat, a system for automatically generating data parallel implementations of Unix shell commands and pipelines. The generated parallel versions split input streams, execute multiple instantiations of the original pipeline commands to process the splits in parallel, then combine the resulting parallel outputs to produce the final output stream.

Yixuan Sun, Chengyao Li, Qian Zhang, Aimin Zhou

In recent years, the prevalence of several pulmonary diseases, especially the coronavirus disease 2019 (COVID-19) pandemic, has attracted worldwide attention. These diseases can be effectively diagnosed and treated with the help of lung imaging. With the development of deep learning technology and the emergence of many public medical image datasets, the diag

Erfan Pakdamanian, Shili Sheng, Sonia Baee, Seongkook Heo

Automated vehicles promise a future where drivers can engage in non-driving tasks without hands on the steering wheels for a prolonged period. Nevertheless, automated vehicles may still need to occasionally hand the control back to drivers due to technology limitations and legal requirements. While some systems determine the need for driver takeover using dr

Boris N. Oreshkin

Adaptive algorithms belong to an important class of algorithms used in radar target detection to overcome prior uncertainty of interference covariance. The contamination of the empirical covariance matrix by the useful signal leads to significant degradation of performance of this class of adaptive algorithms. Regularization, also known in radar literature a

Can Peng, Kun Zhao, Sam Maksoud, Meng Li

Incremental learning requires a model to continually learn new tasks from streaming data. However, traditional fine-tuning of a well-trained deep neural network on a new task will dramatically degrade performance on the old task -- a problem known as catastrophic forgetting. In this paper, we address this issue in the context of anchor-free object detection,

Jacob Nelson, Ahmed Hassan, Roberto Palmieri

We present bundled references, a new building block to provide linearizable range query operations for highly concurrent linked data structures. Bundled references allow range queries to traverse a path through the data structure that is consistent with the target atomic snapshot and is made of the minimal amount of nodes that should be accessed to preserve

Xiaoxian Tang, Zhishuo Zhang

For the reaction networks with one-dimensional stoichiometric subspaces, we show the following results. (1) If the maximum number of positive steady states is an even number N, then the maximum number of stable positive steady states is N/2. (2) If the maximum number of positive steady states is an odd number N, then we provide a condition on the network suc