arXiv subjects

Yuxuan Chen

Publications and source records attributed to Yuxuan Chen.

At least 145 records · Page 8Linked to original sources

arXiv2024

Multigrid on unstructured meshes with regions of low quality cells

The convergence of multigrid methods degrades significantly if a small number of low quality cells are present in a finite element mesh, and this can be a barrier to the efficient and robust application of multigrid on complicated geometric domains. The degraded performance is observed also if intermediate levels in a non-nested geometric multigrid problem have low quality cells, even when the fine grid is high quality. It is demonstrated for geometric multigrid methods that the poor convergence is due to the local failure of smoothers to eliminate parts of error around cells of low quality. To overcome this, a global--local combined smoother is developed to maintain effective relaxation in the presence of a small number of poor quality cells. The smoother involves the application of a standard smoother on the whole domain, followed by local corrections for small subdomains with low quality cells. Two- and three-dimensional numerical experiments demonstrate that the degraded convergence of multigrid for low quality meshes can be restored to the high quality mesh reference case using the proposed smoother. The effect is particularly pronounced for higher-order finite elements. The results provide a basis for developing efficient, non-nested geometric multigrid methods for complicated engineering geometries.

cs.CE↗

arXiv2023

AutoTask: Executing Arbitrary Voice Commands by Exploring and Learning from Mobile GUI

Voice command interfaces (VCIs) have gained increasing importance, enabling hands-free and eyes-free interaction with digital devices. However, the inherent complexity in constructing effective voice interfaces has limited the VCIs' functionalities to only a small fraction of GUI applications and tasks. This paper presents AutoTask, a VCI capable of automating any task in any mobile application without configuration or modification from developers or end users. The primary challenge for AutoTask is the lack of knowledge, as it needs to accomplish unknown tasks (e.g., user commands) within an unknown environment (e.g., GUI). To address this challenge, AutoTask employs two strategies: (1) trial and error: AutoTask explores the GUI, attempts potential operation sequences, and recovers from errors through backtracking; (2) learning from the environment: AutoTask accumulates experiences during exploration and summarizes correct knowledge from these experiences. We implemented AutoTask on Android devices and conducted an evaluation study, which proved the feasibility of AutoTask.

cs.HC↗

arXiv2023

Reconstruction of rough surfaces from a single receiver at grazing angle

The paper develops a method for recovering a one-dimensional rough surface profile from scattered wave field, using a single receiver and repeated measurements when the surface is moving with respect to source and receiver. This extends a previously introduced marching method utilizing low grazing angles, and addresses the key issue of the requirement for many simultaneous receivers. The algorithm recovers the surface height below the receiver point step-by-step as the surface is moved, using the parabolic wave integral equations. Numerical examples of reconstructed surfaces demonstrate that the method is robust in both Dirichlet and Neumann boundary conditions, and with respect to different roughness characteristics and to some degree of measurement noise.

physics.comp-ph↗

arXiv2023

What to Learn: Features, Image Transformations, or Both?

Long-term visual localization is an essential problem in robotics and computer vision, but remains challenging due to the environmental appearance changes caused by lighting and seasons. While many existing works have attempted to solve it by directly learning invariant sparse keypoints and descriptors to match scenes, these approaches still struggle with adverse appearance changes. Recent developments in image transformations such as neural style transfer have emerged as an alternative to address such appearance gaps. In this work, we propose to combine an image transformation network and a feature-learning network to improve long-term localization performance. Given night-to-day image pairs, the image transformation network transforms the night images into day-like conditions prior to feature matching; the feature network learns to detect keypoint locations with their associated descriptor values, which can be passed to a classical pose estimator to compute the relative poses. We conducted various experiments to examine the effectiveness of combining style transfer and feature learning and its training strategy, showing that such a combination greatly improves long-term localization performance.

cs.RO↗

arXiv2023

Towards the Universal Defense for Query-Based Audio Adversarial Attacks

Recently, studies show that deep learning-based automatic speech recognition (ASR) systems are vulnerable to adversarial examples (AEs), which add a small amount of noise to the original audio examples. These AE attacks pose new challenges to deep learning security and have raised significant concerns about deploying ASR systems and devices. The existing defense methods are either limited in application or only defend on results, but not on process. In this work, we propose a novel method to infer the adversary intent and discover audio adversarial examples based on the AEs generation process. The insight of this method is based on the observation: many existing audio AE attacks utilize query-based methods, which means the adversary must send continuous and similar queries to target ASR models during the audio AE generation process. Inspired by this observation, We propose a memory mechanism by adopting audio fingerprint technology to analyze the similarity of the current query with a certain length of memory query. Thus, we can identify when a sequence of queries appears to be suspectable to generate audio AEs. Through extensive evaluation on four state-of-the-art audio AE attacks, we demonstrate that on average our defense identify the adversary intent with over 90% accuracy. With careful regard for robustness evaluations, we also analyze our proposed defense and its strength to withstand two adaptive attacks. Finally, our scheme is available out-of-the-box and directly compatible with any ensemble of ASR defense models to uncover audio AE attacks effectively without model retraining.

eess.AS↗

arXiv2023

Towards the Transferable Audio Adversarial Attack via Ensemble Methods

In recent years, deep learning (DL) models have achieved significant progress in many domains, such as autonomous driving, facial recognition, and speech recognition. However, the vulnerability of deep learning models to adversarial attacks has raised serious concerns in the community because of their insufficient robustness and generalization. Also, transferable attacks have become a prominent method for black-box attacks. In this work, we explore the potential factors that impact adversarial examples (AEs) transferability in DL-based speech recognition. We also discuss the vulnerability of different DL systems and the irregular nature of decision boundaries. Our results show a remarkable difference in the transferability of AEs between speech and images, with the data relevance being low in images but opposite in speech recognition. Motivated by dropout-based ensemble approaches, we propose random gradient ensembles and dynamic gradient-weighted ensembles, and we evaluate the impact of ensembles on the transferability of AEs. The results show that the AEs created by both approaches are valid for transfer to the black box API.

cs.CR↗

arXiv2023

Introduction of an intriguing approach for eletric current transformer on-site examining repairing

The working principle of a electric current transformer is based on electromagnetic induction, mainly composed of a closed iron core and windings. Its primary winding has relatively few turns and is connected in series with the current circuit to be measured. However, due to the frequent occurrence of full current passing through the current transformer during use, its secondary winding has relatively more turns. During use, errors may occur in the current transformer. Therefore, technical personnel can adopt the heterodyne measurement method to shield interference signals and ensure the accuracy of measurements during on-site examining and repairing of current transformers. This article mainly introduces the characteristics of transformers and their on-site examining and repairing process and errors, and describes the specific application of heterodyne measurement in on-site examining and repairing of current transformers.

eess.SY↗

arXiv2023

Structural phase transitions and superconductivity in the Heusler intermetallics $X$Pd$_2$Sn ($X$ = Ti, Zr, Hf)

We report the discovery of structural phase transitions and superconductivity in the full Heusler compounds $X$Pd$_2$Sn ($X$ = Ti, Zr, Hf), by means of electrical transport, magnetic susceptibility, specific heat and x-ray diffraction measurements. TiPd$_2$Sn, ZrPd$_2$Sn and HfPd$_2$Sn undergo structural phase transitions from the room-temperature cubic MnCu$_2$Al-type structure (space group $Fm\bar{3}m$) to a low-temperature tetragonal structure at around 160 K, 110 K and 90 K, respectively, which are likely related charge density wave (CDW) instabilities. Low temperature single crystal x-ray diffraction measurements of ZrPd$_2$Sn demonstrate the emergence of a superstructure with multiple commensurate modulations below $T_s$. ZrPd$_2$Sn and HfPd$_2$Sn have bulk superconductivity (SC) with transition temperatures $T_c$ $\sim$ 1.2 K and 1.3 K, respectively. Density functional theory (DFT) calculations reveal evidence for structural and electronic instabilities which can give rise to CDW formation, suggesting that these $X$Pd$_2$Sn systems are good candidates for examining the interplay between CDW and SC.

cond-mat.supr-con↗

arXiv2022

LGRASS: Linear Graph Spectral Sparsification for Final Task of the 3rd ACM-China International Parallel Computing Challenge

This paper presents our solution for optimization task of the 3rd ACM-China IPCC. By the complexity analysis, we identified three time-consuming subroutines of original algorithm: marking edges, computing pseudo inverse and sorting edges. These subroutines becomes the main performance bottleneck owing to their super-linear time complexity. To address this, we proposed LGRASS, a linear graph spectral sparsification algorithm to run in strictly linear time. LGRASS takes advantage of spanning tree properties and efficient algorithms to optimize bottleneck subroutines. Furthermore, we crafted a parallel processing scheme for LGRASS to make full use of multi-processor hardware. Experiment shows that our proposed method fulfils the task in dozens of milliseconds on official test cases and keep its linearity as graph size scales up on random test cases.

cs.DC↗

arXiv2022

Self-Supervised Feature Learning for Long-Term Metric Visual Localization

Visual localization is the task of estimating camera pose in a known scene, which is an essential problem in robotics and computer vision. However, long-term visual localization is still a challenge due to the environmental appearance changes caused by lighting and seasons. While techniques exist to address appearance changes using neural networks, these methods typically require ground-truth pose information to generate accurate image correspondences or act as a supervisory signal during training. In this paper, we present a novel self-supervised feature learning framework for metric visual localization. We use a sequence-based image matching algorithm across different sequences of images (i.e., experiences) to generate image correspondences without ground-truth labels. We can then sample image pairs to train a deep neural network that learns sparse features with associated descriptors and scores without ground-truth pose supervision. The learned features can be used together with a classical pose estimator for visual stereo localization. We validate the learned features by integrating with an existing Visual Teach & Repeat pipeline to perform closed-loop localization experiments under different lighting conditions for a total of 22.4 km.

cs.RO↗

arXiv2022

Multilingual Relation Classification via Efficient and Effective Prompting

Prompting pre-trained language models has achieved impressive performance on various NLP tasks, especially in low data regimes. Despite the success of prompting in monolingual settings, applying prompt-based methods in multilingual scenarios has been limited to a narrow set of tasks, due to the high cost of handcrafting multilingual prompts. In this paper, we present the first work on prompt-based multilingual relation classification (RC), by introducing an efficient and effective method that constructs prompts from relation triples and involves only minimal translation for the class labels. We evaluate its performance in fully supervised, few-shot and zero-shot scenarios, and analyze its effectiveness across 14 languages, prompt variants, and English-task training in cross-lingual settings. We find that in both fully supervised and few-shot scenarios, our prompt method beats competitive baselines: fine-tuning XLM-R_EM and null prompts. It also outperforms the random baseline by a large margin in zero-shot experiments. Our method requires little in-language knowledge and can be used as a strong baseline for similar multilingual classification tasks.

cs.CL↗

arXiv2022

Creating a Nanoscale Lateral Heterojunction in a Semiconductor Monolayer with a Large Built-in Potential

The ability to engineer atomically thin nanoscale lateral heterojunctions (HJs) is critical to lay the foundation for future two-dimensional (2D) device technology. However, the traditional approach to creating a heterojunction by direct growth of a heterostructure of two different materials constrains the available band offsets, and it is still unclear if large built-in potentials are attainable for 2D materials. The electronic properties of atomically thin semiconducting transition metal dichalcogenides (TMDs) are not static, and their exciton binding energy and quasiparticle band gap depend strongly on the proximal environment. Recent studies have shown that this effect can be harnessed to engineer the lateral band profile of monolayer TMDs to create a heterojunction. Here we demonstrate the synthesis of a nanoscale lateral heterojunction in monolayer MoSe2 by intercalating Se at the interface of a hBN/Ru(0001) substrate. The Se intercalation creates a spatially abrupt modulation of the local hBN/Ru work function, which is imprinted directly onto an overlying MoSe2 monolayer to create a large built-in potential of 0.83 eV. We spatially resolve the MoSe2 band profile and work function using scanning tunneling spectroscopy to map out the nanoscale depletion region. The Se intercalation also modifies the dielectric environment, influencing the local band gap renormalization and increasing the MoSe2 band gap by ~0.26 eV. This work illustrates that environmental proximity engineering provides a robust method to indirectly manipulate the band profile of 2D materials outside the limits of their intrinsic properties, providing avenues for future device design.

cond-mat.mtrl-sci↗

arXiv2022

SLIC: Self-Supervised Learning with Iterative Clustering for Human Action Videos

Self-supervised methods have significantly closed the gap with end-to-end supervised learning for image classification. In the case of human action videos, however, where both appearance and motion are significant factors of variation, this gap remains significant. One of the key reasons for this is that sampling pairs of similar video clips, a required step for many self-supervised contrastive learning methods, is currently done conservatively to avoid false positives. A typical assumption is that similar clips only occur temporally close within a single video, leading to insufficient examples of motion similarity. To mitigate this, we propose SLIC, a clustering-based self-supervised contrastive learning method for human action videos. Our key contribution is that we improve upon the traditional intra-video positive sampling by using iterative clustering to group similar video instances. This enables our method to leverage pseudo-labels from the cluster assignments to sample harder positives and negatives. SLIC outperforms state-of-the-art video retrieval baselines by +15.4% on top-1 recall on UCF101 and by +5.7% when directly transferred to HMDB51. With end-to-end finetuning for action classification, SLIC achieves 83.2% top-1 accuracy (+0.8%) on UCF101 and 54.5% on HMDB51 (+1.6%). SLIC is also competitive with the state-of-the-art in action classification after self-supervised pretraining on Kinetics400.

cs.CV↗

arXiv2022

Why only Micro-F1? Class Weighting of Measures for Relation Classification

Relation classification models are conventionally evaluated using only a single measure, e.g., micro-F1, macro-F1 or AUC. In this work, we analyze weighting schemes, such as micro and macro, for imbalanced datasets. We introduce a framework for weighting schemes, where existing schemes are extremes, and two new intermediate schemes. We show that reporting results of different weighting schemes better highlights strengths and weaknesses of a model.

cs.CL↗

arXiv2022

A Comparative Study of Pre-trained Encoders for Low-Resource Named Entity Recognition

Pre-trained language models (PLM) are effective components of few-shot named entity recognition (NER) approaches when augmented with continued pre-training on task-specific out-of-domain data or fine-tuning on in-domain data. However, their performance in low-resource scenarios, where such data is not available, remains an open question. We introduce an encoder evaluation framework, and use it to systematically compare the performance of state-of-the-art pre-trained representations on the task of low-resource NER. We analyze a wide range of encoders pre-trained with different strategies, model architectures, intermediate-task fine-tuning, and contrastive learning. Our experimental results across ten benchmark NER datasets in English and German show that encoder performance varies significantly, suggesting that the choice of encoder for a specific low-resource scenario needs to be carefully evaluated.

cs.CL↗

arXiv2021

PASTO: Strategic Parameter Optimization in Recommendation Systems -- Probabilistic is Better than Deterministic

Real-world recommendation systems often consist of two phases. In the first phase, multiple predictive models produce the probability of different immediate user actions. In the second phase, these predictions are aggregated according to a set of 'strategic parameters' to meet a diverse set of business goals, such as longer user engagement, higher revenue potential, or more community/network interactions. In addition to building accurate predictive models, it is also crucial to optimize this set of 'strategic parameters' so that primary goals are optimized while secondary guardrails are not hurt. In this setting with multiple and constrained goals, this paper discovers that a probabilistic strategic parameter regime can achieve better value compared to the standard regime of finding a single deterministic parameter. The new probabilistic regime is to learn the best distribution over strategic parameter choices and sample one strategic parameter from the distribution when each user visits the platform. To pursue the optimal probabilistic solution, we formulate the problem into a stochastic compositional optimization problem, in which the unbiased stochastic gradient is unavailable. Our approach is applied in a popular social network platform with hundreds of millions of daily users and achieves +0.22% lift of user engagement in a recommendation task and +1.7% lift in revenue in an advertising optimization scenario comparing to using the best deterministic parameter strategy.

cs.LG↗

arXiv2021

SoK: A Modularized Approach to Study the Security of Automatic Speech Recognition Systems

With the wide use of Automatic Speech Recognition (ASR) in applications such as human machine interaction, simultaneous interpretation, audio transcription, etc., its security protection becomes increasingly important. Although recent studies have brought to light the weaknesses of popular ASR systems that enable out-of-band signal attack, adversarial attack, etc., and further proposed various remedies (signal smoothing, adversarial training, etc.), a systematic understanding of ASR security (both attacks and defenses) is still missing, especially on how realistic such threats are and how general existing protection could be. In this paper, we present our systematization of knowledge for ASR security and provide a comprehensive taxonomy for existing work based on a modularized workflow. More importantly, we align the research in this domain with that on security in Image Recognition System (IRS), which has been extensively studied, using the domain knowledge in the latter to help understand where we stand in the former. Generally, both IRS and ASR are perceptual systems. Their similarities allow us to systematically study existing literature in ASR security based on the spectrum of attacks and defense solutions proposed for IRS, and pinpoint the directions of more advanced attacks and the directions potentially leading to more effective protection in ASR. In contrast, their differences, especially the complexity of ASR compared with IRS, help us learn unique challenges and opportunities in ASR security. Particularly, our experimental study shows that transfer learning across ASR models is feasible, even in the absence of knowledge about models (even their types) and training data.

cs.CR↗

arXiv2021

Boosting Few-Shot Classification with View-Learnable Contrastive Learning

The goal of few-shot classification is to classify new categories with few labeled examples within each class. Nowadays, the excellent performance in handling few-shot classification problems is shown by metric-based meta-learning methods. However, it is very hard for previous methods to discriminate the fine-grained sub-categories in the embedding space without fine-grained labels. This may lead to unsatisfactory generalization to fine-grained subcategories, and thus affects model interpretation. To tackle this problem, we introduce the contrastive loss into few-shot classification for learning latent fine-grained structure in the embedding space. Furthermore, to overcome the drawbacks of random image transformation used in current contrastive learning in producing noisy and inaccurate image pairs (i.e., views), we develop a learning-to-learn algorithm to automatically generate different views of the same image. Extensive experiments on standard few-shot learning benchmarks demonstrate the superiority of our method.

cs.CV↗