October 2022 arXiv papers — page 114

Ali Zafari, Atefeh Khoshkhahtinat, Piyush M. Mehta, Nasser M. Nasrabadi

NASA's Solar Dynamics Observatory (SDO) mission gathers 1.4 terabytes of data each day from its geosynchronous orbit in space. SDO data includes images of the Sun captured at different wavelengths, with the primary scientific goal of understanding the dynamic processes governing the Sun. Recently, end-to-end optimized artificial neural networks (ANN) have sh

Mohsen Mesgar, Thy Thy Tran, Goran Glavas, Iryna Gurevych

Few-shot Intent Classification (FSIC) is one of the key challenges in modular task-oriented dialog systems. While advanced FSIC methods are similar in using pretrained language models to encode texts and nearest neighbour-based inference for classification, these methods differ in details. They start from different pretrained text encoders, use different enc

Yuesong Shen, Daniel Cremers

Neural networks are known to produce poor uncertainty estimations, and a variety of approaches have been proposed to remedy this issue. This includes deep ensemble, a simple and effective method that achieves state-of-the-art results for uncertainty-aware learning tasks. In this work, we explore a combinatorial generalization of deep ensemble called deep com

Jonathan Scott, Michelle Yeo, Christoph H. Lampert

We present Cross-Client Label Propagation(XCLP), a new method for transductive federated learning. XCLP estimates a data graph jointly from the data of multiple clients and computes labels for the unlabeled data by propagating label information across the graph. To avoid clients having to share their data with anyone, XCLP employs two cryptographically secur

Nikhil Parthasarathy, S. M. Ali Eslami, João Carreira, Olivier J. Hénaff

Humans learn powerful representations of objects and scenes by observing how they evolve over time. Yet, outside of specific tasks that require explicit temporal understanding, static image pretraining remains the dominant paradigm for learning visual foundation models. We question this mismatch, and ask whether video pretraining can yield visual representat

Bing Huang, O. Anatole von Lilienfeld, Jaron T. Krogel, Anouar Benali

In the past decade, quantum diffusion Monte Carlo (DMC) has been demonstrated to successfully predict the energetics and properties of a wide range of molecules and solids by numerically solving the electronic many-body Schr\"odinger equation. We show that when coupled with quantum machine learning (QML) based surrogate methods the computational burden can b

David Benisty, N. Wyn Evans, Anne-Christine Davis

Future Gaia and Legacy Survey of Space and Time data releases, together with wide area spectroscopic surveys, will deliver large samples of resolved binary stars with phase space coordinates, albeit with low-cadence. Given an eccentricity law $f(\epsilon)$, we derive properties of (i) the velocity distribution $v/\sqrt{G M/r}$ normalised by the value for a c

Haotao Wang, Junyuan Hong, Aston Zhang, Jiayu Zhou

Deep neural networks (DNNs) are vulnerable to backdoor attacks. Previous works have shown it extremely challenging to unlearn the undesired backdoor behavior from the network, since the entire network can be affected by the backdoor samples. In this paper, we propose a brand-new backdoor defense strategy, which makes it much easier to remove the harmful infl

Conor Mullaney, Adnan Aijaz, Nathan Sealey, Ben Holden

Peer-to-Peer (P2P) energy trading provides various benefits over conventional wholesale energy markets and makes renewable energy more accessible. This paper proposes a novel multi-layer P2P energy trading system for microgrids based on IOTA 2.0, which is a distributed ledger technology (DLT) primarily designed for Internet-of-Things (IoT) applications. The

Juan Carlos Forte

This work reports the detection of a multi peaked colour pattern in the integrated colours distribution of globular clusters associated to the giant elliptical galaxy NGC 4486, using Next Generation Virgo Survey data. This feature is imprinted on the well known bimodal colour distribution of these clusters. Remarkably, the pattern is similar to that found in

Mohammadmahdi Nouriborji, Omid Rohanian, Samaneh Kouchaki, David A. Clifton

Pre-trained Language Models (LMs) have become an integral part of Natural Language Processing (NLP) in recent years, due to their superior performance in downstream applications. In spite of this resounding success, the usability of LMs is constrained by computational and time complexity, along with their increasing size; an issue that has been referred to a

Abigail Hickok

Persistence diagram (PD) bundles, a generalization of vineyards, were introduced as a way to study the persistent homology of a set of filtrations parameterized by a topological space $B$. In this paper, we present an algorithm for computing piecewise-linear PD bundles, a wide class that includes many of the PD bundles that one may encounter in practice. Ful

Hongyu Wang, Shuming Ma, Shaohan Huang, Li Dong

A big convergence of model architectures across language, vision, speech, and multimodal is emerging. However, under the same name "Transformers", the above areas use different implementations for better performance, e.g., Post-LayerNorm for BERT, and Pre-LayerNorm for GPT and vision Transformers. We call for the development of Foundation Transformer for tru

Fredrik Hellström, Giuseppe Durisi

We present a new family of information-theoretic generalization bounds, in which the training loss and the population loss are compared through a jointly convex function. This function is upper-bounded in terms of the disintegrated, samplewise, evaluated conditional mutual information (CMI), an information measure that depends on the losses incurred by the s

Carlos Desa, Weyner Ccuiro, David Choque

We present a family of $AdS_{2+1}$ asymptotically hairy black holes, within the context of Einstein's minimally coupled scalar field theory. We study the boundary conditions and construct the thermal superpotential. In the Euclidean section, we calculate the free energy and in the Lorentzian section, the Brown-York tensor, both regularized by two methods. Fi

Gregory S. Adkins, Jonathan Gomprecht, Yanxi Li, Evan Shinn

We have completed the calculation of pure-recoil corrections of order $(Z \alpha)^6$ to Coulombic bound states of two spin-1/2 fermions without approximation in the particle masses. Our result applies to systems of arbitrary mass ratio such as muonium and positronium, and also hydrogen and muonic hydrogen (with the neglect of proton structure effects). We ha

Andrew M. Childs, Robin Kothari, Matt Kovacs-Deak, Aarthi Sundaram

The divide-and-conquer framework, used extensively in classical algorithm design, recursively breaks a problem of size $n$ into smaller subproblems (say, $a$ copies of size $n/b$ each), along with some auxiliary work of cost $C^{\textrm{aux}}(n)$, to give a recurrence relation $$C(n) \leq a \, C(n/b) + C^{\textrm{aux}}(n)$$ for the classical complexity $C(n)

Ieva Staliūnaitė, Philip John Gorinski, Ignacio Iacobacci

Multihop Question Answering is a complex Natural Language Processing task that requires multiple steps of reasoning to find the correct answer to a given question. Previous research has explored the use of models based on Graph Neural Networks for tackling this task. Various architectures have been proposed, including Relational Graph Convolutional Networks

Agapi Rissaki, Bruno Scarone, David Liu, Aditeya Pandey

The issue of bias (i.e., systematic unfairness) in machine learning models has recently attracted the attention of both researchers and practitioners. For the graph mining community in particular, an important goal toward algorithmic fairness is to detect and mitigate bias incorporated into graph embeddings since they are commonly used in human-centered appl

Amit Kachroo, Sai Prashanth Chinnapalli

The application of machine learning (ML) techniques in wireless communication domain has seen a tremendous growth over the years especially in the wireless sensing domain. However, the questions surrounding the ML model's inference reliability, and uncertainty associated with its predictions are never answered or communicated properly. This itself raises a l

Yu Du, Gabriel Kosmacher, Yichen Liu, Jeff Massman

Exploiting the relationship between 4-dimensional toric and semitoric integrable systems with Delzant and semitoric polygons, respectively, we develop techniques to compute certain equivariant packing densities and equivariant capacities of these systems by working exclusively with the polygons. This expands on results of Pelayo and Pelayo-Schmidt. We comput

Félix del Teso, Jørgen Endal, Espen R. Jakobsen, Juan Luis Vázquez

We consider the evolution problem associated to the infinity fractional Laplacian introduced by Bjorland, Caffarelli and Figalli (2012) as the infinitesimal generator of a non-Brownian tug-of-war game. We first construct a class of viscosity solutions of the initial-value problem for bounded and uniformly continuous data. An important result is the equivalen

Jason Phang, Herbie Bradley, Leo Gao, Louis Castricato

Over the past two years, EleutherAI has established itself as a radically novel initiative aimed at both promoting open-source research and conducting research in a transparent, openly accessible and collaborative manner. EleutherAI's approach to research goes beyond transparency: by doing research entirely in public, anyone in the world can observe and cont

Valerie Domcke, Kohei Kamada, Kyohei Mukaida, Kai Schmitz

CP violation and the violation of baryon-minus-lepton number B-L do not necessarily have to occur simultaneously in order to accomplish successful leptogenesis. Instead, it suffices if new CP-violating interactions at high energies result in primordial charge asymmetries, which are then reprocessed into a nonvanishing B-L asymmetry by right-handed neutrinos

Tom Rindell, Berat Yenilen, Niklas Halonen, Arttu Pönni

We study the approximate state preparation problem on noisy intermediate-scale quantum (NISQ) computers by applying a genetic algorithm to generate quantum circuits for state preparation. The algorithm can account for the specific characteristics of the physical machine in the evaluation of circuits, such as the native gate set and qubit connectivity. We use

Jonathan Pham, Sofia Wyetzner, Martin R. Pfaller, David W. Parker

We propose svMorph, a framework for interactive virtual sculpting of patient-specific vascular anatomic models. Our framework includes three tools for the creation of tortuosity, aneurysms, and stenoses in tubular vascular geometries. These shape edits are performed via geometric operations on the surface mesh and vessel centerline curves of the input model.

Shirin Panahi, Matteo Lodi, Marco Storace, Francesco Sorrentino

In this paper, we study the network pinning control problem in the presence of two different types of coupling: (i) node-to-node coupling among the network nodes and (ii) input-to-node coupling from the source node to the `pinned nodes'. Previous work has mainly focused on the case that (i) and (ii) are of the same type. We decouple the stability analysis of

Shaobo Lin, Xingyu Zeng, Rui Zhao

Conventional training of deep neural networks usually requires a substantial amount of data with expensive human annotations. In this paper, we utilize the idea of meta-learning to explain two very different streams of few-shot learning, i.e., the episodic meta-learning-based and pre-train finetune-based few-shot learning, and form a unified meta-learning fr

Ishan Jindal, Alexandre Rademaker, Khoi-Nguyen Tran, Huaiyu Zhu

Semantic role labeling (SRL) identifies the predicate-argument structure in a sentence. This task is usually accomplished in four steps: predicate identification, predicate sense disambiguation, argument identification, and argument classification. Errors introduced at one step propagate to later steps. Unfortunately, the existing SRL evaluation scripts do n

Corey Lynch, Ayzaan Wahid, Jonathan Tompson, Tianli Ding

We present a framework for building interactive, real-time, natural language-instructable robots in the real world, and we open source related assets (dataset, environment, benchmark, and policies). Trained with behavioral cloning on a dataset of hundreds of thousands of language-annotated trajectories, a produced policy can proficiently execute an order of

Giacomo Del Nin, Raquel Perales

We prove that given an $n$-dimensional integral current space and a $1$-Lipschitz map, from this space onto the $n$-dimensional Euclidean ball, that preserves the mass of the current and is injective on the boundary, then the map has to be an isometry. We deduce as a consequence a stability result with respect to the intrinsic flat distance, which implies th

Md Sakib Ullah Sourav, Huidong Wang, Mohammad Sultan Mahmud, Hua Zheng

In this research, we propose a complete set of approaches for identifying and extracting emotions from Bangla texts. We provide a Bangla emotion classifier for six classes: anger, disgust, fear, joy, sadness, and surprise, from Bangla words using transformer-based models, which exhibit phenomenal results in recent days, especially for high-resource languages

Tong Liu, Hadi Meidani

Rapid reliability assessment of transportation networks can enhance preparedness, risk mitigation, and response management procedures related to these systems. Network reliability analysis commonly considers network-level performance and does not consider the more detailed node-level responses due to computational cost. In this paper, we propose a rapid seis

Alex Samuel Bamunoba, Innocent Ndikubwayo

Given coprime integers $k, \ell$ with $k > \ell \geqslant 1$ and arbitrary complex polynomials $A(z), B(z)$ with $\deg(A(z)B(z))\geqslant 1$, we consider the polynomial sequence $\{P_n(z)\}$ satisfying a three-term recurrence $P_n(z)+B(z)P_{n-\ell}(z)+A(z)P_{n-k}(z)=0$ subject to the initial conditions $P_0(z)=1$, $P_{-1}(z)=\cdots=P_{1-k}(z)=0$ and fully ch

Anna Kh. Balci, Lars Diening, Johannes Storn

We introduce a globally convergent relaxed Kacanov scheme for the computation of the discrete minimizer to the $p$-Laplace problem with $2 \leq p < \infty$. The iterative scheme is easy to implement since each iterate results only from the solve of a weighted, linear Poisson problem. It neither requires an additional line search nor involves unknown constant

Zhipeng Cai, Vladlen Koltun, Ozan Sener

Given a stream of data sampled from non-stationary distributions, online continual learning (OCL) aims to adapt efficiently to new data while retaining existing knowledge. The typical approach to address information retention (the ability to retain previous knowledge) is keeping a replay buffer of a fixed size and computing gradients using a mixture of new d

Fan Feng, Chen Liang, Dongdong Chen, Ke Du

The space-variant wavefront reconstruction problem inherently exists in deep tissue imaging. In this paper,we propose a framework of Shack-Hartmann wavefront space-variant sensing with extended source illumination. The space-variant wavefront is modeled as a four-dimensional function where two dimensionsare in the spatial domain and two in the Fourier domain

M. H. Modirrousta, M. Aliyari Shoorehdeli, M. Yari, A. Ghahremani

In modern industrial systems, diagnosing faults in time and using the best methods becomes more and more crucial. It is possible to fail a system or to waste resources if faults are not detected or are detected late. Machine learning and deep learning have proposed various methods for data-based fault diagnosis, and we are looking for the most reliable and p

Anna Balaudo, Alice Garoffolo, Matteo Martinelli, Suvodip Mukherjee

We investigate the synergy of upcoming galaxy surveys and gravitational wave (GW) experiments in constraining late-time cosmology, examining the cross-correlations between the weak lensing of gravitational waves (GW-WL) and the galaxy fields. Without focusing on any specific GW detector configuration, we benchmark the requirements for the high-precision meas

Prithul Sarker, Khondker Fariha Hossain, Isayas Berhe Adhanom, Philip K Pavilionis

The sport-related concussion (SRC) battery relies heavily upon subjective symptom reporting in order to determine the diagnosis of a concussion. Unfortunately, athletes with SRC may return-to-play (RTP) too soon if they are untruthful of their symptoms. It is critical to provide accurate assessments that can overcome underreporting to prevent further injury.

Himanshu Kumar Sudhanshu, Sudhaker Upadhyay, Dharm Veer Singh, Sunil Kumar

We study the corrected entropy due to thermal fluctuation and their effect on the thermodynamics of a conformally dressed black hole in three dimensions. We find that the thermal fluctuation affects the entropy significantly for small black holes. Various corrected thermodynamical variables are also calculated for this black hole. We observe that thermal flu

Jamie Bamber

I discuss how one can apply the covariant formalism developed by Vilkovisky and DeWitt to obtain frame invariant fifth force calculations for scalar-tensor theories. Fifth forces are severely constrained by astrophysical measurements. It was shown previously that for scale-invariant Higgs-dilaton gravity, in a particular choice of Jordan frame, the dilaton f

Fabio Ciolli, Francesco Fidaleo

For spectral actions consisting of the average number of particles and arising from open systems made of general free $q$-particles (including Bose, Fermi and classical ones corresponding to $q=\pm 1$ and $0$, respectively) in thermal equilibrium, we compute the asymptotic expansion with respect to the natural cut-off. We treat both relevant situations relat

Adam Wolsey

Artificial Intelligence techniques have evolved rapidly in recent years, revolutionising the approaches used to fight against cybercriminals. But as the cyber security field has progressed, so has malware development, making it an economic imperative to strengthen businesses' defensive capability against malware attacks. This review aims to outline the state

Sharan Narasimhan, Pooja Shekar, Suvodip Dey, Maunendra Sankar Desarkar

Text Style Transfer (TST) is performable through approaches such as latent space disentanglement, cycle-consistency losses, prototype editing etc. The prototype editing approach, which is known to be quite successful in TST, involves two key phases a) Masking of source style-associated tokens and b) Reconstruction of this source-style masked sentence conditi

Mustafa Can Çavdar, Ibrahim Korpeoglu, Özgür Ulusoy

Wireless sensor networks (WSNs) have many applications and are an essential part of IoT systems. The primary functionality of a WSN is gathering data from specific points that are covered with sensor nodes and transmitting the collected data to remote units for further processing. In IoT use cases, a WSN infrastructure may need to be shared by many applicati

Hunter Rehm, Mona Matar, Puck Rombach, Lauren McIntyre

Katz centrality is a popular network centrality measure. It takes a (weighted) count of all walks starting at each node, with an additional damping factor of $\alpha$ that tunes the influence of walks as lengths increase. We introduce a tool to compare different centrality measures in terms of their node rankings, which takes into account that a relative ran

Hans Hao-Hsun Hsu, Yuesong Shen, Christian Tomani, Daniel Cremers

Given the importance of getting calibrated predictions and reliable uncertainty estimations, various post-hoc calibration methods have been developed for neural networks on standard multi-class classification tasks. However, these methods are not well suited for calibrating graph neural networks (GNNs), which presents unique challenges such as accounting for

Matthias Samland, Timothy Brandt, Julien Milli, Philippe Delorme

We present a new open-source data-reduction pipeline to reconstruct spectral data cubes from raw SPHERE integral-field spectrograph (IFS) data. The pipeline is written in Python and based on the pipeline that was developed for the CHARIS IFS. It introduces several improvements to SPHERE data analysis that ultimately produce significant improvements in postpr

Lorena López-Hernanz, Rudy Rosas

We prove a two-dimensional analog of Leau-Fatou flower theorem for non-degenerate reduced tangent to the identity biholomorphisms.

Bradley Jenks, Filippo Pecci, Ivan Stoianov

The provision of self-cleaning velocities has been shown to reduce the risk of discolouration in water distribution networks (WDNs). Despite these findings, control implementations continue to be focused primarily on pressure and leakage management. This paper considers the control of diurnal flow velocities to maximize the self-cleaning capacity (SCC) of WD

Bálint Farkas, Béla Nagy, Szilárd Gy. Révész

In previous papers we investigated so-called sum of translates functions $F({\mathbf{x}},t):=J(t)+\sum_{j=1}^n \nu_j K(t-x_j)$, where $J:[0,1]\to \underline{\mathbb{R}}:={\mathbb{R}}\cup\{-\infty\}$ is a "sufficiently nondegenerate" and upper-bounded "field function", and $K:[-1,1]\to \underline{\mathbb{R}}$ is a fixed "kernel function", concave both on $(-1

Lang Feng, Qianhui Liu, Huajin Tang, De Ma

Spiking neural networks (SNNs) are bio-inspired neural networks with asynchronous discrete and sparse characteristics, which have increasingly manifested their superiority in low energy consumption. Recent research is devoted to utilizing spatio-temporal information to directly train SNNs by backpropagation. However, the binary and non-differentiable propert

Shuo Wang, Chen Qin, Chengyan Wang, Kang Wang

The quality of cardiac magnetic resonance (CMR) imaging is susceptible to respiratory motion artifacts. The model robustness of automated segmentation techniques in face of real-world respiratory motion artifacts is unclear. This manuscript describes the design of extreme cardiac MRI analysis challenge under respiratory motion (CMRxMotion Challenge). The cha

Eldar Kurtic, Dan Alistarh

We revisit the performance of the classic gradual magnitude pruning (GMP) baseline for large language models, focusing on the classic BERT benchmark on various popular tasks. Despite existing evidence in the literature that GMP performs poorly, we show that a simple and general variant, which we call GMP*, can match and sometimes outperform more complex stat

Sebastian Ohrem, Wolfgang Reichel, Roland Schnaubelt

We consider the linear wave equation $V(x) u_{tt}(x, t) - u_{xx}(x, t) = 0$ on $[0, \infty)\times[0, \infty)$ with initial conditions and a nonlinear Neumann boundary condition $u_x(0, t) = (f(u_t(0,t)))_t$ at $x=0$. This problem is an exact reduction of a nonlinear Maxwell problem in electrodynamics. In the case where $f\colon\mathbb{R}\to\mathbb{R}$ is an

Chao-Han Huck Yang, Jun Qi, Sabato Marco Siniscalchi, Chin-Hui Lee

We propose an ensemble learning framework with Poisson sub-sampling to effectively train a collection of teacher models to issue some differential privacy (DP) guarantee for training data. Through boosting under DP, a student model derived from the training data suffers little model degradation from the models trained with no privacy protection. Our proposed

Marios Constantinides, Daniele Quercia

Pervasive technologies combined with powerful AI have been recently introduced to enhance work productivity. Yet, some of these technologies are judged to be invasive. To identify which ones, we should understand how employees tend to judge these technologies. We considered 16 technologies that track productivity, and conducted a study in which 131 crowd-wor

Manish Prajapat, Matteo Turchetta, Melanie N. Zeilinger, Andreas Krause

In multi-agent coverage control problems, agents navigate their environment to reach locations that maximize the coverage of some density. In practice, the density is rarely known $\textit{a priori}$, further complicating the original NP-hard problem. Moreover, in many applications, agents cannot visit arbitrary locations due to $\textit{a priori}$ unknown s

Gregor Geigle, Chen Cecilia Liu, Jonas Pfeiffer, Iryna Gurevych

Current multimodal models, aimed at solving Vision and Language (V+L) tasks, predominantly repurpose Vision Encoders (VE) as feature extractors. While many VEs -- of different architectures, trained on different data and objectives -- are publicly available, they are not designed for the downstream V+L tasks. Nonetheless, most current work assumes that a \te

Zachary Munro

We generalize ideas of Jahncke from trees to square complexes. We introduce the notion of progression in $\mathrm{CAT}(0)$ square complexes. Using progression, we are able to build on the proof strategy of Dahmani-Guirardel-Przytycki to show any action of a random group with seven or more generators on a $\mathrm{CAT}(0)$ square complex has a global fixed po

Sirui Song, Kirk Saunders, Ye Yue, Jundong Liu

Collision avoidance is a crucial task in vision-guided autonomous navigation. Solutions based on deep reinforcement learning (DRL) has become increasingly popular. In this work, we proposed several novel agent state and reward function designs to tackle two critical issues in DRL-based navigation solutions: 1) smoothness of the trained flight trajectories; a

Daniel Loureiro, Alípio Mário Jorge

Progress on commonsense reasoning is usually measured from performance improvements on Question Answering tasks designed to require commonsense knowledge. However, fine-tuning large Language Models (LMs) on these specific tasks does not directly evaluate commonsense learned during pre-training. The most direct assessments of commonsense knowledge in pre-trai

Caleb Koch, Carmen Strassle, Li-Yang Tan

We establish new hardness results for decision tree optimization problems, adding to a line of work that dates back to Hyafil and Rivest in 1976. We prove, under randomized ETH, superpolynomial lower bounds for two basic problems: given an explicit representation of a function $f$ and a generator for a distribution $\mathcal{D}$, construct a small decision t

Sohaib Khalid, Zakarias Sjöström Dyrefelt

We show that on any compact K\"ahler surface existence of solutions to the Z-critical equation can be characterized using a finite number of effective conditions, where the number of conditions is bounded above by the Picard number of the surface.This leads to a first PDE analogue of the locally finite wall-chamber decomposition in Bridgeland stability. As a

Lei Li, Nicolas Donati, Maks Ovsjanikov

In this work, we present a novel non-rigid shape matching framework based on multi-resolution functional maps with spectral attention. Existing functional map learning methods all rely on the critical choice of the spectral resolution hyperparameter, which can severely affect the overall accuracy or lead to overfitting, if not chosen carefully. In this paper

Jianqiao Liu, Xu-Ping Yao, Gang Chen

Like the twisting control in magic angle twisted bilayer graphenes, the stacking control is another mechanical approach to manipulate the fundamental properties of solids, especially the van der Waals materials. We explore the stacking-induced magnetic frustration and the spiral spin liquid on a multilayer triangular lattice antiferromagnet where the system

Peng Song, Zhufeng Hou, Kenta Hongo, Ryo Maezono

The recent high-pressure experimental discovery of superconductivity in (La,Y)H$_{10}$, (La,Ce)H$_{9}$, (La,Ce)H$_{10}$, (Y,Ce)H$_{9}$, and (La,Nd)H$_{10}$ shows that the ternary rare-earth clathrate hydride can be promising candidate for high-temperature superconductor. In this work, we theoretically demonstrate that the combination of actinide-metal thoriu

Sewade Ogun, Vincent Colotte, Emmanuel Vincent

Training of multi-speaker text-to-speech (TTS) systems relies on curated datasets based on high-quality recordings or audiobooks. Such datasets often lack speaker diversity and are expensive to collect. As an alternative, recent studies have leveraged the availability of large, crowdsourced automatic speech recognition (ASR) datasets. A major problem with su

Alexandre Martin

We show that two-dimensional Artin groups satisfy a strengthening of the Tits alternative: their subgroups either contain a non-abelian free group or are virtually free abelian of rank at most $2$. When in addition the associated Coxeter group is hyperbolic, we answer in the affirmative a question of Wise on the subgroups generated by large powers of two ele

Tao Sun, Nidal Abuhajar, Shuyu Gong, Zhewei Wang

Speaker separation aims to extract multiple voices from a mixed signal. In this paper, we propose two speaker-aware designs to improve the existing speaker separation solutions. The first model is a speaker conditioning network that integrates speech samples to generate individualized speaker conditions, which then provide informed guidance for a separation

Baptiste Goujaud, Adrien Taylor, Aymeric Dieuleveut

In this work, we propose an adaptive variation on the classical Heavy-ball method for convex quadratic minimization. The adaptivity crucially relies on so-called "Polyak step-sizes", which consists in using the knowledge of the optimal value of the optimization problem at hand instead of problem parameters such as a few eigenvalues of the Hessian of the prob

Ting Chen, Lala Li, Saurabh Saxena, Geoffrey Hinton

Panoptic segmentation assigns semantic and instance ID labels to every pixel of an image. As permutations of instance IDs are also valid solutions, the task requires learning of high-dimensional one-to-many mapping. As a result, state-of-the-art approaches use customized architectures and task-specific loss functions. We formulate panoptic segmentation as a

P. R. N. Falcão, M. L. Lyra

We consider a disordered one-dimensional tight-binding model with power-law decaying hopping amplitudes to disclose wavefunction maximum distributions related to the Anderson localization phenomenon. Deeply in the regime of extended states, the wavefunction intensities follow the Porter-Thomas distribution while their maxima assume the Gumbel distribution. A

Shiv Ram Dubey, Satish Kumar Singh, Bidyut Baran Chaudhuri

The stochastic gradient descent (SGD) optimizers are generally used to train the convolutional neural networks (CNNs). In recent years, several adaptive momentum based SGD optimizers have been introduced, such as Adam, diffGrad, Radam and AdaBelief. However, the existing SGD optimizers do not exploit the gradient norm of past iterations and lead to poor conv

Zhou Li, Hua Sun

A storage code over a graph maps $K$ independent source symbols, each of $L_w$ bits, to $N$ coded symbols, each of $L_v$ bits, such that each coded symbol is stored in a node of the graph and each edge of the graph is associated with one source symbol. From a pair of nodes connected by an edge, the source symbol that is associated with the edge can be decode

Binhua Liao, Yani Chen, Zhewei Wang, Charles D. Smith

In this paper, we explore the capabilities of a number of deep neural network models in generating whole-brain 3T-like MR images from clinical 1.5T MRIs. The models include a fully convolutional network (FCN) method and three state-of-the-art super-resolution solutions, ESPCN [26], SRGAN [17] and PRSR [7]. The FCN solution, U-Convert-Net, carries out mapping

Sadasivan Shankar, Albert Reuther

We examine the computational energy requirements of different systems driven by the geometrical scaling law, and increasing use of Artificial Intelligence or Machine Learning (AI-ML) over the last decade. With more scientific and technology applications based on data-driven discovery, machine learning methods, especially deep neural networks, have become wid

Dehua Zheng, Xiaochen Zheng, Laurence T. Yang, Yuan Gao

Recent research about camouflaged object detection (COD) aims to segment highly concealed objects hidden in complex surroundings. The tiny, fuzzy camouflaged objects result in visually indistinguishable properties. However, current single-view COD detectors are sensitive to background distractors. Therefore, blurred boundaries and variable shapes of the camo

Veronica Felli, Giulio Romani

We study singular perturbations of eigenvalues of the polyharmonic operator on bounded domains under removal of small interior compact sets. We consider both homogeneous Dirichlet and Navier conditions on the external boundary, while we impose homogeneous Dirichlet conditions on the boundary of the removed set. To this aim, we develop a notion of capacity wh

H. Miao, T. T. Zhang, H. X. Li, G. Fabbris

The intertwining between spin, charge, and lattice degrees of freedom can give rise to unusual macroscopic quantum states, including high-temperature superconductivity and quantum anomalous Hall effects. Recently, a charge density wave (CDW) is observed in the kagome antiferromagnet FeGe, indicative of possible intertwining physics. An outstanding question i

François Lamoline, Anthony Hastir

Stochastic infinite-dimensional port-Hamiltonian systems (SPHSs) with multiplicative Gaussian white noise are considered. In this article we extend the notion of Dirac structure for deterministic distributed parameter port-Hamiltonian systems to a stochastic ones by adding some additional stochastic ports. Using the Stratonovich formalism of the stochastic i

Sumeet Kulkarni, Surendra Padamata, Anuradha Gupta

The LIGO-Virgo gravitational wave detectors have confidently observed 4 events involving neutron stars: two binary neutron star (BNS) mergers (GW170817 and GW190425), and two neutron star-black hole mergers (GW200105 and GW200115). However, our theoretical understanding of the remnant properties of such systems is incomplete due to the complexities related t

Róbert Csordás, Kazuki Irie, Jürgen Schmidhuber

Well-designed diagnostic tasks have played a key role in studying the failure of neural nets (NNs) to generalize systematically. Famous examples include SCAN and Compositional Table Lookup (CTL). Here we introduce CTL++, a new diagnostic dataset based on compositions of unary symbolic functions. While the original CTL is used to test length generalization or

Dan Su, Mostofa Patwary, Shrimai Prabhumoye, Peng Xu

Closed-book question answering (QA) requires a model to directly answer an open-domain question without access to any external knowledge. Prior work on closed-book QA either directly finetunes or prompts a pretrained language model (LM) to leverage the stored knowledge. However, they do not fully exploit the parameterized knowledge. To address this issue, we

Ishwor Poudyal, Zhi Qiao, Arndt Last, Michael R. Armstrong

A pump-probe dark-field X-ray microscopy (DFXM) experiment was carried out at the Advanced Photon Source (APS) with nanosecond drive laser pulses and hybrid mode X-ray probe pulses. We observe a thermal decay due to laser-induced heat diffusion in a Germanium single crystal which matches the theoretical prediction. The single pulse DFXM imaging in combinatio

Xiangyu Luo, Hao Chen, Yinghao Li, Qiang Gao

In high temperature cuprate superconductors, it was found that the superconducting transition temperature Tc depends on the number of CuO2 planes (n) in the structural unit and the maximum Tc is realized in the trilayer system (n=3). It was also found that the trilayer superconductors exhibit an unusual phase diagram that Tc keeps nearly constant in the over

Emanuele Dolera, Enrico Priola

Let $(\lambda_k)$ be a strictly increasing sequence of positive numbers such that $\sum_{k=1}^{\infty} \frac{1}{\lambda_k} < \infty.$ Let $f $ be a bounded smooth function and denote by $u= u^f$ the bounded classical solution to $u(x) - \frac{1}{2}\sum_{k=1}^m D^2_{kk} u(x) + \sum_{k =1}^m \lambda_k x_k D_k u(x) = f(x), $ $ x \in \R^m$. It is known that the

Chenji Li, Brato Chakrabarti, Pedro Castilla, Achal Mahajan

The propulsion of mammalian spermatozoa relies on the spontaneous periodic oscillation of their flagella. These oscillations are driven internally by the coordinated action of ATP-powered dynein motors that exert sliding forces between microtubule doublets, resulting in bending waves that propagate along the flagellum and enable locomotion. We present an int

Ojas Kishorkumar Shirekar, Anuj Singh, Hadi Jamali-Rad

Humans have a unique ability to learn new representations from just a handful of examples with little to no supervision. Deep learning models, however, require an abundance of data and supervision to perform at a satisfactory level. Unsupervised few-shot learning (U-FSL) is the pursuit of bridging this gap between machines and humans. Inspired by the capacit

Hao Li, Bo-Qiang Ma

It has been reported that the Large High Altitude Air Shower Observatory (LHAASO) observed very high energy photons from GRB 221009A, with the highest energy reaching 18~TeV. We find that observation of such high energy photons is quite nontrivial since extragalactic background light could absorb these photons severely and the flux is too weak to be observed

Shenyang Tan, Wenjun Liu

In order to find a better physical model to describe the large-scale cloud-water transformation and rainfall, we consider a moist atmosphere model consisting of the primitive equations with only horizontal viscosity in the dynamic equation and a set of humidity equations describing water vapor, rain water and cloud condensates. To overcome difficulties cause

Max M. Chumley, Melih C. Yesilli, Jisheng Chen, Firas A. Khasawneh

Quantifying patterns in visual or tactile textures provides important information about the process or phenomena that generated these patterns. In manufacturing, these patterns can be intentionally introduced as a design feature, or they can be a byproduct of a specific process. Since surface texture has significant impact on the mechanical properties and th

B. J. Lynch, B. E. Wood, M. Jin, T. Török

The aim of this white paper is to briefly summarize some of the outstanding gaps in the observations and modeling of stellar flares, CMEs, and exoplanetary space weather, and to discuss how the theoretical and computational tools and methods that have been developed in heliophysics can play a critical role in meeting these challenges. The maturity of data-in

Mai Nishimura, Shohei Nobuhara, Ko Nishino

We introduce a novel learning-based method for view birdification, the task of recovering ground-plane trajectories of pedestrians of a crowd and their observer in the same crowd just from the observed ego-centric video. View birdification becomes essential for mobile robot navigation and localization in dense crowds where the static background is hard to se

Somin Wadhwa, Vivek Khetan, Silvio Amir, Byron Wallace

We present Reddit Health Online Talk (RedHOT), a corpus of 22,000 richly annotated social media posts from Reddit spanning 24 health conditions. Annotations include demarcations of spans corresponding to medical claims, personal experiences, and questions. We collect additional granular annotations on identified claims. Specifically, we mark snippets that de

Xiaojian Xu, Weijie Gan, Satya V. V. N. Kothapalli, Dmitriy A. Yablonskiy

Quantitative MRI (qMRI) refers to a class of MRI methods for quantifying the spatial distribution of biological tissue parameters. Traditional qMRI methods usually deal separately with artifacts arising from accelerated data acquisition, involuntary physical motion, and magnetic-field inhomogeneities, leading to suboptimal end-to-end performance. This paper

Wei Wang, Ting Zhang

In this paper, we consider the homogenization problem for generalized elliptic systems $$ \mathcal{L}_{\varepsilon}=-\operatorname{div}(A(x/\varepsilon)\nabla+V(x/\varepsilon))+B(x/\varepsilon)\nabla+c(x/\varepsilon)+λI $$ with dimension two. Precisely, we will establish the $ W^{1,p} $ estimates, Hölder estimates, Lipschitz estimates and $ L^p $ convergence

Hoang Tran, Ashok Cutkosky

We introduce new algorithms and convergence guarantees for privacy-preserving non-convex Empirical Risk Minimization (ERM) on smooth $d$-dimensional objectives. We develop an improved sensitivity analysis of stochastic gradient descent on smooth objectives that exploits the recurrence of examples in different epochs. By combining this new approach with recen

George Peters, Diogo Pacheco

This paper aims to reduce randomness in football by analysing the role of lineups in final scores using machine learning prediction models we have developed. Football clubs invest millions of dollars on lineups and knowing how individual statistics translate to better outcomes can optimise investments. Moreover, sports betting is growing exponentially and be