检索结果-内蒙古大学图书馆

2022 Findings of the Association for Computational Linguistics: EMNLP 2022

作者： Zhu, Jiahao Liu, Daizong Zhou, Pan Di, Xing Cheng, Yu Yang, Song Xu, Wenzheng Xu, Zichuan Wan, Yao Sun, Lichao Xiong, Zeyu Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Huazhong University of Science and Technology China ProtagoLabs Inc Microsoft Research Wangxuan Institute of Computer Technology Peking University China Beijing Institute of Technology China School of Sichuan University China Dalian University of Technology China Lehigh University United States School of Computer Sci. & Tech. Huazhong University of Science and Technology China

Temporal sentence grounding (TSG) aims to identify the temporal boundary of a specific segment from an untrimmed video by a sentence query. All existing works first utilize a sparse sampling strategy to extract a fixed number of video frames and then conduct multimodal interactions with query sentence for reasoning. However, we argue that these methods have overlooked two indispensable issues: 1) Boundary-bias: The annotated target segment generally refers to two specific frames as corresponding start and end timestamps. The video downsampling process may lose these two frames and take the adjacent irrelevant frames as new boundaries. 2) Reasoning-bias: Such incorrect new boundary frames also lead to the reasoning bias during frame-query interaction, reducing the generalization ability of model. To alleviate above limitations, in this paper, we propose a novel Siamese Sampling and Reasoning Network (SSRN) for TSG, which introduces a siamese sampling mechanism to generate additional contextual frames to enrich and refine the new boundaries. Specifically, a reasoning strategy is developed to learn the inter-relationship among these frames and generate soft labels on boundaries for more accurate frame-query reasoning. Such mechanism is also able to supplement the absent consecutive visual semantics to the sampled sparse frames for fine-grained activity understanding. Extensive experiments demonstrate the effectiveness of SSRN on three challenging datasets. © 2022 Association for Computational Linguistics.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

POViT: Vision Transformer for Multi-objective Design and Characterization of Nanophotonic Devices

arXiv

引用

arXiv 2022年

作者： Chen, Xinyu Li, Renjie Yu, Yueyao Shen, Yuanwen Li, Wenye Zhang, Zhaoyu Zhang, Yin School of Science and Engineering The Chinese University of Hong Kong Shenzhen China Shenzhen Key Laboratory of Semiconductor Lasers China School of Data Science The Chinese University of Hong Kong Shenzhen China

We solve a fundamental challenge in semiconductor IC design: the fast and accurate characterization of nanoscale photonic devices. Much like the fusion between AI and EDA, many efforts have been made to apply DNNs such as convolutional neural networks (CNN) to prototype and characterize next-gen optoelectronic devices commonly found in photonic integrated circuits (PICs). These prior works generally strive to predict the quality factor (Q) and modal volume (V) of for instance, photonic crystals, with ultra-high accuracy and speed. However, state-of-the-art models are still far from being directly applicable in the real-world: e.g. the correlation coefficient of V (Vcoeff) is only about 80%, which is much lower than what it takes to generate reliable and reproducible nanophotonic designs. Recently, attention-based transformer models have attracted extensive interests and been widely used in CV and NLP. In this work, we propose the first-ever Transformer model (POViT) to efficiently design and simulate semiconductor photonic devices with multiple objectives. Unlike the standard Vision Transformer (ViT), we supplied photonic crystals as data input and changed the activation layer from GELU [1] to an absolute-value function (ABS). Our experiments show that POViT exceeds results reported by previous models significantly. The correlation coefficient Vcoeffincreases by over 12% (i.e., to 92:0%) and the prediction errors of Q is reduced by an order of magnitude, among several other key metric improvements. Our work has the potential to drive the expansion of EDA to fully automated photonic design. The complete dataset and code will be released to aid researchers endeavoring in the interdisciplinary field of physics and computer science. © 2022, CC BY-NC-SA.

关键词： Photonic devices

来源：评论

学校读者我要写书评

暂无评论

Downstream-agnostic Adversarial Examples

arXiv

引用

arXiv 2023年

作者： Zhou, Ziqi Hu, Shengshan Zhao, Ruizhi Wang, Qian Zhang, Leo Yu Hou, Junhui Jin, Hai School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Computer Science and Technology Huazhong University of Science and Technology China School of Cyber Science and Engineering Wuhan University China School of Information and Communication Technology Griffith University Australia Department of Computer Science City University of Hong Kong Hong Kong National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security China Hubei Engineering Research Center on Big Data Security China Cluster and Grid Computing Lab

Self-supervised learning usually uses a large amount of unlabeled data to pre-train an encoder which can be used as a general-purpose feature extractor, such that downstream users only need to perform fine-tuning operations to enjoy the benefit of "large model". Despite this promising prospect, the security of pre-trained encoder has not been thoroughly investigated yet, especially when the pre-trained encoder is publicly available for commercial use. In this paper, we propose AdvEncoder, the first framework for generating downstream-agnostic universal adversarial examples based on the pre-trained encoder. AdvEncoder aims to construct a universal adversarial perturbation or patch for a set of natural images that can fool all the downstream tasks inheriting the victim pre-trained encoder. Unlike traditional adversarial example works, the pre-trained encoder only outputs feature vectors rather than classification labels. Therefore, we first exploit the high frequency component information of the image to guide the generation of adversarial examples. Then we design a generative attack framework to construct adversarial perturbations/patches by learning the distribution of the attack surrogate dataset to improve their attack success rates and transferability. Our results show that an attacker can successfully attack downstream tasks without knowing either the pre-training dataset or the downstream dataset. We also tailor four defenses for pre-trained encoders, the results of which further prove the attack ability of AdvEncoder. Our codes are available at: https://github. com/CGCL-codes/AdvEncoder. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Once Read is Enough: Domain-specific Pretraining-free Language Models with Cluster-guided Sparse Experts for Long-tail Domain knowledge 38

Once Read is Enough: Domain-specific Pretraining-free Langua...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Dong, Fang Chen, Mengyi Zhou, Jixian Shi, Yubin Chen, Yixuan Dong, Mingzhi Wang, Yujiang Li, Dongsheng Yang, Xiaochen Zhu, Rui Dick, Robert Lv, Qin Yang, Fan Lu, Tun Gu, Ning Shang, Li China and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China Oxford Suzhou Centre for Advanced Research Suzhou China University of Bath United Kingdom Microsoft Research Asia Shanghai China School of Mathematics & Statistics University of Glasgow United Kingdom Bayes Business School City St George's University of London United Kingdom Department of Electrical Engineering and Computer Science University of Michigan United States Department of Computer Science University of Colorado Boulder United States School of Microelectronics Fudan University China

Language models (LMs) only pretrained on a general and massive corpus usually cannot attain satisfying performance on domain-specific downstream tasks, and hence, applying domain-specific pretraining to LMs is a common and indispensable practice. However, domain-specific pretraining can be costly and time-consuming, hindering LMs' deployment in real-world applications. In this work, we consider the incapability to memorize domain-specific knowledge embedded in the general corpus with rare occurrences and "long-tail" distributions as the leading cause for pretrained LMs' inferior downstream performance. Analysis of Neural Tangent Kernels (NTKs) reveals that those long-tail data are commonly overlooked in the model's gradient updates and, consequently, are not effectively memorized, leading to poor domain-specific downstream performance. Based on the intuition that data with similar semantic meaning are closer in the embedding space, we devise a Cluster-guided Sparse Expert (CSE) layer to actively learn long-tail domain knowledge typically neglected in previous pretrained LMs. During pretraining, a CSE layer efficiently clusters domain knowledge together and assigns long-tail knowledge to designate extra experts. CSE is also a lightweight structure that only needs to be incorporated in several deep layers. With our training strategy, we found that during pretraining, data of long-tail knowledge gradually formulate isolated, "outlier" clusters in an LM's representation spaces, especially in deeper layers. Our experimental results show that only pretraining CSE-based LMs is enough to achieve superior performance than regularly pretrained-finetuned LMs on various downstream tasks, implying the prospects of domain-specific-pretraining-free language models. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Substitutional Doped GeSe: Tunable Oxidative States with Strain engineering

arXiv

引用

arXiv 2022年

作者： Shu, Zheng Cai, Yongqing Center for Data Science Institute of Collaborative Innovation University of Macau Taipa China Joint Key Laboratory The Ministry of Education Institute of Applied Physics and Materials Engineering University of Macau Taipa China

Layered chalcogenide materials have a wealth of nanoelectronics applications like resistive switching and energy-harvesting such as photocatalyst owing to rich electronic, orbital, and lattice excitations. In this work, we explore monochalcogenide germanium selenide GeSe with respect to substitutional doping with 13 metallic cations by using first-principles calculations. Typical dopants including s-shell (alkali elements Li and Na), p-shell (Al, Pb and Bi), 3d (Fe, Cu, Co and Ni), 4d (Pd and Ag) and 5d (Au and Pt) elements are systematically examined. Amongst all the cationic dopants, Al with the highest oxidation states, implying a high mobility driven by electric field, and Al-doped GeSe may be a promising candidate for novel resistive switching devices. We show that there exist many localized induced states in the band gap of GeSe upon doping Fe, Co, or Ni, while for Cu, Ag, and Au cases there is no such states in the gap. The Ag and Cu are + 0.27 and + 0.35 charged respectively and the positive charges are beneficial for field-driven motion in GeSe. In contrast, Au is slightly negatively charged renders Au-doped GeSe a promising photocatalyst and enhanced surface plasmon. Moreover, we explore the coexistence of dopant and strain in GeSe and find dynamical adjustments of localized states in GeSe with levels successive shifting upward/downward with strain. This induces dynamic oxidative states of the dopants under strain which should be quite popular in composites where motion of metal adatoms causes significant deformation. © 2022, CC BY-SA.

关键词： Germanium compounds

来源：评论

学校读者我要写书评

暂无评论

Structured Tensor Decomposition Based Channel Estimation and Double Refinements for Active RIS Empowered Broadband Systems

arXiv

引用

arXiv 2024年

作者： Wang, Yirun Wang, Yongqing Shen, Yuyao Wang, Gongpu Tellambura, Chintha School of Information and Electronics Beijing Institute of Technology Beijing100081 China Beijing Key Laboratory of Transportation Data Analysis and Mining Beijing Jiaotong University Beijing100044 China Department of Electrical and Computer Engineering University of Alberta EdmontonABT6G 2R3 Canada

Channel parameter recovery is critical for the next-generation reconfigurable intelligent surface (RIS)-empowered communications and sensing. Tensor-based mechanisms are particularly effective, inherently capturing the multi-dimensional nature of wireless channels. However, existing studies assume either a line-of-sight (LOS) scenario or a blocked TX-RX channel. This paper solves a novel problem: tensor-based channel parameter estimation for active RIS-aided multiple-antenna broadband connections in fully multipath environments with the TX-RX link. System settings are customized to construct a fifth-order canonical polyadic (CP) signal tensor that matches the five-dimensional channel. Four tensor factors contain redundant columns, rendering the classical Kruskal’s condition for decomposition uniqueness unsatisfied. The fifth-order Vandermonde structured CP decomposition (VSCPD) is developed to address this challenge, making the tensor factorization problem solvable using only linear algebra and offering a relaxed general uniqueness condition. With VSCPD as a perfect decoupling scheme, a sequential triple-stage channel estimation algorithm is proposed based on one-dimensional parameter estimation. The first stage enables multipath identification and algebraic coarse estimation. The following two stages offer optional successive refinements at the cost of increased complexity. The closed-form Cramér-Rao lower bound (CRLB) is derived to assess the estimation performance. Herein, the noise covariance matrix depends on multipath parameters in our active-RIS scenario. Numerical results are provided to verify the effectiveness of proposed algorithms under various evaluation metrics. Our results also show that active RIS can significantly improve channel estimation performance compared to passive RIS. Copyright © 2024, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

Activation of Phosphorene-like Two-dimensional GeSe for Efficient Electrocatalytic Nitrogen Reduction via States Filtering of Ru

arXiv

引用

arXiv 2022年

Nitrogen reduction reaction (NRR) which converts nitrogen (N2) to ammonia (NH3) normally requires harsh conditions to break the bound nitrogen bond. Herein, via first-principles calculation we reveal that a superior NRR catalytic activity could be obtained through anchoring atomic catalyst above a phosphorene-like puckering surface of germanium selenide (GeSe). Through examining the single- and double- atoms (B, Fe, W, Mo and Ru) decorated on GeSe, we find that its rippled structure allows an intimate contact between the deposited species and the GeSe which significantly promotes the states hybridization. Amongst the various atomic catalyst, we predict that the Ru dimer decorated GeSe monolayer (Ru2@GeSe) has superior catalytic activity for the N2 fixation and reduction. Through examining the three NRR pathways (distal, alternating and enzymatic), the distal and enzymatic pathway is both the thermodynamically favorable with the maximum Gibbs free energy change (ΔGMAX) of 0.25 and 0.26 eV, respectively. Such a superior activity could be attributed to the filtered states of GeSe by Ru dimer which leads to the effective activation of the adsorbed N2 bond. As an efficient near-infrared absorber of GeSe, the Ru mediated hybridization of GeSe-RuN2 complex enables an in-gap state which further broadens the absorbing window, rendering for a broadband solar absorption and possible photocatalysis. © 2022, CC BY-SA.

关键词： Atoms

来源：评论

学校读者我要写书评

暂无评论

GANTEE: Generative Adversatial Network for Taxonomy Entering Evaluation

arXiv

引用

arXiv 2023年

作者： Gu, Zhouhong Jiang, Sihang Liu, Jingping Xiao, Yanghua Feng, Hongwei Li, Zhixu Liang, Jiaqing Zhong, Jian Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Information Science and Engineering East China University of Science and Technology China Fudan-Aishu Cognitive Intelligence Joint Research Center China School of Data Science Fudan University China HUAWEI CBG Edu AI Lab. China

Taxonomy is formulated as directed acyclic concepts graphs or trees that support many downstream tasks. Many new coming concepts need to be added to an existing taxonomy. The traditional taxonomy expansion task aims only at finding the best position for new coming concepts in the existing taxonomy. However, they have two drawbacks when being applied to the real-scenarios. The previous methods suffer from low-efficiency since they waste much time when most of the new coming concepts are indeed noisy concepts. They also suffer from low-effectiveness since they collect training samples only from the existing taxonomy, which limits the ability of the model to mine more hypernym-hyponym relationships among real concepts. This paper proposes a pluggable framework called Generative Adversarial Network for Taxonomy Entering Evaluation (GANTEE) to alleviate these drawbacks. A generative adversarial network is designed in this framework by discriminative models to alleviate the first drawback and the generative model to alleviate the second drawback. Two discriminators are used in GANTEE to provide long-term and short-term rewards, respectively. Moreover, to further improve the efficiency, pre-trained language models are used to retrieve the representation of the concepts quickly. The experiments on three real-world large-scale datasets with two different languages show that GANTEE improves the performance of the existing taxonomy expansion methods in both effectiveness and efficiency. © 2023, CC BY.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

A Class of Convex Optimization-Based Recursive Algorithms for Identification of Stochastic Systems

arXiv

引用

arXiv 2024年

作者： Ding, Mingxia Zhao, Wenxiao Chen, Tianshi Key Laboratory of Systems and Control Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing100190 China School of Mathematical Sciences University of Chinese Academy of Sciences Beijing100049 China School of Science and Engineering Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen518172 China

Focusing on identification, this paper develops a class of convex optimization-based criteria and correspondingly the recursive algorithms to estimate the parameter vector θ*of a stochastic dynamic system. Not only do the criteria include the classical least-squares estimator but also the Ll = | |l, l ≥ 1, the Huber, the Logcosh, and the Quantile costs as special cases. First, we prove that the minimizers of the convex optimization-based criteria converge to θ*with probability one. Second, the recursive algorithms are proposed to find the estimates, which minimize the convex optimization-based criteria, and it is shown that these estimates also converge to the true parameter vector with probability one. Numerical examples are given, justifying the performance of the proposed algorithms including the strong consistency of the estimates, the robustness against outliers in the observations, and higher efficiency in online computation compared with the kernel-based regularization method due to the recursive nature. Copyright © 2024, The Authors. All rights reserved.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Online Public Transit Ridership Flow Estimation through Passive WiFi Sensing

Online Public Transit Ridership Flow Estimation through Pass...

引用

第34届中国控制与决策会议

作者： Wenbo Chang Baoqi Huang Bing Jia Wuyungerile Li Gang Xu Engineering Research Center of Ecological Big Data Ministry of Education Inner Mongolia A.R.Key Laboratory of Wireless Networking and Mobile Computing College of Computer Science Inner Mongolia University

Online public transit ridership flow information is helpful to improve the service quality of urban public transportation and travel experience of *** WiFi sensing collects WiFi probe requests sent by mobile devices in a non-intrusive manner,and can thus be used to monitor ridership *** with the existing methods based on manual configurations or video surveillance,passive WiFi sensing based methods demonstrate the advantages of small interference,large coverage,low cost and simple *** recent years,researchers proposed some offline methods based on passive WiFi sensing,but offering online information is still *** paper develops a public transit ridership flow monitoring system using customized WiFi sniffers,and proposes an algorithm to implement online public transit ridership flow ***,the proposed algorithm uses the bidirectional long short-term memory neural network(BiLSTM) to capture the bidirectional time series features contained in WiFi sensing data,and further integrates the attention mechanism to focus on the critical features in relation to *** experimental results,carried out on four bus routes in Hohhot,China,show that the RMSE and MAE errors of the proposed algorithm are 3.96 and 3.1,respectively,which are better than those of the other 7 commonly used online ***,the accuracy of the proposed algorithm is 5% lower than that of the corresponding offline algorithm,but it is evident that our online algorithm will stimulate more practical applications.

关键词： Passive WiFi sensing Online Public transit ridership flow Attention mechanism BiLSTM

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：