检索结果-内蒙古大学图书馆

TRUST: Triangle counting reloaded on GPUs

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Pandey, Santosh Wang, Zhibin Zhong, Sheng Tian, Chen Zheng, Bolong Li, Xiaoye Li, Lingda Hoisie, Adolfy Ding, Caiwen Li, Dong Liu, Hang Electrical and Computer Engineering Stevens Institute of Technology United States State Key Laboratory for Novel Software Technology Nanjing University Nanjing Jiangsu210008 China Huazhong University of Science and Technology China Computational Research Division Lawrence Berkeley National Laboratory United States Brookhaven National Laboratory United States Department of Computer Science & Engineering University of Connecticut United States Department of Electrical Engineering and Computer Science University of California Merced United States

Triangle counting is a building block for a wide range of graph applications. Traditional wisdom suggests that i) hashing is not suitable for triangle counting, ii) edge-centric triangle counting beats vertex-centric design, and iii) communication-free and workload balanced graph partitioning is a grand challenge for triangle counting. On the contrary, we advocate that i) hashing can help the key operations for scalable triangle counting on Graphics Processing Units (GPUs), i.e., list intersection and graph partitioning, ii) vertex-centric design reduces both hash table construction cost and memory consumption, which is limited on GPUs. In addition, iii) we exploit graph and workload collaborative, and hashing-based 2D partitioning to scale vertex-centric triangle counting over 1,000 GPUs with sustained scalability. In this work, we present TRUST which performs triangle counting with the hash operation and vertex-centric mechanism at the core. To the best of our knowledge, TRUST is the first work that achieves over one trillion Traversed Edges Per Second (TEPS) rate for triangle counting. © 2021, CC BY.

关键词： Program processors

Multi-Sources Fusion Learning for Multi-Points NLOS Localization in OFDM System

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Bohao Shuai, Zitao Huang, Chongwen Yang, Qianqian Yang, Zhaohui Jin, Richeng Hammadi, Ahmed Al Zhang, Zhaoyang Yuen, Chau Debbah, Mérouane The College of Information Science and Electronic Engineering Zhejiang University Hangzhou310007 China The State Key Laboratory of Integrated Service Networks Xidian University Xi’an710071 China Hangzhou310027 China The Department of Electrical Engineering and Computer Science University of Michigan Ann ArborMI48109-2122 United States The Technology Innovation Institute Masdar City Abu Dhabi9639 United Arab Emirates The School of Electrical and Electronics Engineering Nanyang Technological University Singapore639798 Singapore KU 6G Research Center Khalifa University of Science and Technology PO Box 127788 Abu Dhabi United Arab Emirates

Accurate localization of mobile terminals is a pivotal aspect of integrated sensing and communication systems. Traditional fingerprint-based localization methods, which infer coordinates from channel information within pre-set rectangular areas, often face challenges due to the heterogeneous distribution of fingerprints inherent in non-line-of-sight (NLOS) scenarios, particularly within orthogonal frequency division multiplexing systems. To overcome this limitation, we develop a novel multi-sources information fusion learning framework referred to as the Autosync Multi-Domains NLOS Localization (AMDNLoc). Specifically, AMDNLoc employs a two-stage matched filter fused with a target tracking algorithm and iterative centroid-based clustering to automatically and irregularly segment NLOS regions, ensuring uniform distribution within channel state information across frequency, power, and time-delay domains. Additionally, the framework utilizes a segment-specific linear classifier array, coupled with deep residual network-based feature extraction and fusion, to establish the correlation function between fingerprint features and coordinates within these regions. Simulation results reveal that AMDNLoc achieves an impressive NLOS localization accuracy of 1.46 meters on typical wireless artificial intelligence research datasets and demonstrates significant improvements in interpretability, adaptability, and scalability. Copyright © 2024, The Authors. All rights reserved.

关键词： Orthogonal frequency division multiplexing

Frequency Self-Adaptation Graph Neural Network for Unsupervised Graph Anomaly Detection

学校读者我要写书评

暂无评论

Neural networks : the official journal of the International Neural Network Society 2025年 190卷 107612页

作者： Ming Gu Gaoming Yang Zhuonan Zheng Meihan Liu Haishuai Wang Jiawei Chen Sheng Zhou Jiajun Bu College of Computer Science and Technology Zhejiang University Hangzhou 310027 China. Electronic address: guming444@***. Taobao and Tmall Group Hangzhou 310052 China. Electronic address: yanggaoming.ygm@***. College of Computer Science and Technology Zhejiang University Hangzhou 310027 China. Electronic address: zhengzn@***. College of Computer Science and Technology Zhejiang University Hangzhou 310027 China. Electronic address: lmh_zju@***. College of Computer Science and Technology Zhejiang University Hangzhou 310027 China Zhejiang Key Laboratory of Accessible Perception and Intelligent Systems Zhejiang University Hangzhou 310027 China. Electronic address: haishuai.wang@***. College of Computer Science and Technology Zhejiang University Hangzhou 310027 China. Electronic address: sleepyhunt@***. Zhejiang Key Laboratory of Accessible Perception and Intelligent Systems Zhejiang University Hangzhou 310027 China School of Software Technology Zhejiang University Ningbo 315048 China. Electronic address: zhousheng_zju@***. Zhejiang Key Laboratory of Accessible Perception and Intelligent Systems Zhejiang University Hangzhou 310027 China. Electronic address: bjj@***.

Unsupervised Graph Anomaly Detection (UGAD) seeks to identify abnormal patterns in graphs without relying on labeled data. Among existing UGAD methods, Graph Neural Networks (GNNs) have played a critical role in learning effective representation for detection by filtering low-frequency graph signals. However, the presence of anomalies can shift the frequency band of graph signals toward higher frequencies, thereby violating the fundamental assumptions underlying GNNs and anomaly detection frameworks. To address this challenge, the design of novel graph filters has garnered significant attention, with recent approaches leveraging anomaly labels in a semi-supervised manner. Nonetheless, the absence of anomaly labels in real-world scenarios has rendered these methods impractical, leaving the question of how to design effective filters in an unsupervised manner largely unexplored. To bridge this gap, we propose a novel Frequency Self-Adaptation Graph Neural Network for Unsupervised Graph Anomaly Detection (FAGAD). Specifically, FAGAD adaptively fuses signals across multiple frequency bands using full-pass signals as a reference. It is optimized via a self-supervised learning approach, enabling the generation of effective representations for unsupervised graph anomaly detection. Experimental results demonstrate that FAGAD achieves state-of-the-art performance on both artificially injected datasets and real-world datasets. The code and datasets are publicly available at https://***/eaglelab-zju/FAGAD.

关键词： Graph Neural Networks Graph filtering Unsupervised Graph Anomaly Detection

EC²Detect: Real-Time Online Video Object Detection in Edge-Cloud Collaborative IoT

学校读者我要写书评

暂无评论

IEEE Internet of Things Journal 2022年第20期9卷 20382-20392页

作者： Siyan Guo Cong Zhao Guiqin Wang Jiaqing Yang Shusen Yang National Engineering Laboratory for Big Data Analytics and the School of Mathematics and Statistics Xi’an Jiaotong University Xi’an China Department of Computing Imperial College London London U.K. National Engineering Laboratory for Big Data Analytics and the School of Computer Science and Technology Xi’an Jiaotong University Xi’an China School of Mathematics and Statistics Xi’an Jiaotong University Xi’an China National Engineering Laboratory for Big Data Analytics and Ministry of Education Key Laboratory for Intelligent Networks and Network Security Xi’an Jiaotong University Xi’an China Industrial Artificial Intelligence Center Pazhou Laboratory Guangzhou China

Video object detection is a fundamental technology of intelligent video analytics for Internet of Things (IoT) applications. However, even with extraordinary detection accuracy, predominating solutions based on deep convolutional neural networks (DCNNs) cannot achieve real-time online object detection on video streams with a low end-to-end (E2E) response latency and therefore cannot be applied to proliferating latency-sensitive IoT applications like autonomous driving requiring large-scale intelligent video analytics. To address this issue, we present EC2Detect, an edge-cloud collaborative real-time online video object detection method. Specifically, we propose a tracking-assisted object detection architecture based on edge-cloud collaboration with keyframe selection, where the accurate but heavy object detection is conducted by the Cloud on sparse keyframes adaptively selected according to their semantic variation, and the lightweight object tracking is used to localize and identify objects in other frames at edge devices. Extensive experiments of our real-world prototype demonstrate that, EC2Detect significantly outperforms state-of-the-art methods in terms of processing speed (up to $4.77\times $ faster), E2E latency (up to $8.12 \times $ lower), and edge-cloud bandwidth occupation ( $17 \times $ lower) with an acceptable mAP, which can effectively support large-scale intelligent video analytics in practice. Source code of EC2Detect is available at https://***/ECCDetect/ECCDetect .

关键词： Object detection Image edge detection Streaming media Internet of Things Collaboration Visual analytics Feature extraction

Pro-UIGAN: Progressive Face Hallucination from Occluded Thumbnails

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhang, Yang Yu, Xin Lu, Xiaobo Liu, Ping The Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China The Australian Institute of Artificial Intelligence University of Technology Sydney UltimoNSW2007 Australia The School of Automation Southeast University Nanjing210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China Singapore138634 Singapore

In this paper, we study the task of hallucinating an authentic high-resolution (HR) face from an occluded thumbnail. We propose a multi-stage Progressive Upsampling and Inpainting Generative Adversarial Network, dubbed Pro-UIGAN, which exploits facial geometry priors to replenish and upsample (8×) the occluded and tiny faces (16×16 pixels). Pro-UIGAN iteratively (1) estimates facial geometry priors for low-resolution (LR) faces and (2) acquires non-occluded HR face images under the guidance of the estimated priors. Our multi-stage hallucination network upsamples and inpaints occluded LR faces via a coarse-to-fine fashion, significantly reducing undesirable artifacts and blurriness. Specifically, we design a novel cross-modal attention module for facial priors estimation, in which an input face and its landmark features are formulated as queries and keys, respectively. Such a design encourages joint feature learning across the input facial and landmark features, and deep feature correspondences will be discovered by attention. Thus, facial appearance features and facial geometry priors are learned in a mutually beneficial manner. Extensive experiments show that our Pro-UIGAN attains visually pleasing completed HR faces, thus facilitating downstream tasks, i.e., face alignment, face parsing, face recognition as well as expression classification. Copyright © 2021, The Authors. All rights reserved.

关键词： Generative adversarial networks

Post-storm vehicle routing for distribution grid restoration: An OLUCT based learning approach

学校读者我要写书评

暂无评论

Post-storm vehicle routing for distribution grid restoration...

2020 IEEE Power and Energy Society General Meeting, PESGM 2020

作者： Shuai, Hang He, Haibo Wen, Jinyu University of Rhode Island Department of Electrical Computer and Biomedical Engineering KingstonRI02881 United States Huazhong University of Science and Technology State Key Laboratory of Advanced Electromagnetic Engineering and Technology School of Electrical and Electronic Engineering Wuhan China

ISBN: (纸本)9781728155081

Natural disasters such as storms usually bring significant damages to distribution grids. Many distribution grids are not equipped with sensors that can pinpoint the location of actual faults. This further complicated the outage restoration process in distribution grids. This paper focuses on the optimal routing of repair vehicles to restore outages in the grid as fast as possible after a storm. First, the vehicle routing problem is formulated as a sequential stochastic optimization problem. In the optimization model, the belief state of the power grid is updated according to the phone calls from customers and the information collected by the repair vehicles. Second, an Open Loop Upper Confidence Bound for Trees (OLUCT) algorithm based stochastic lookahead policy is utilized to achieve the realtime dispatching of the repair vehicles. Simulation results show that the proposed approach can effectively navigate the vehicle to repair all outages. © 2020 IEEE.

关键词： Optimization

Self-Prompting Large Language Models for Zero-Shot Open-Domain QA

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Junlong Wang, Jinyuan Zhang, Zhuosheng Zhao, Hai Department of Computer Science and Engineering Shanghai Jiao Tong University China SJTU-Paris Elite Institute of Technology Shanghai Jiao Tong University China School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University China

Open-Domain Question Answering (ODQA) aims to answer questions without explicitly providing specific background documents. This task becomes notably challenging in a zeroshot setting where no data is available to train tailored retrieval-reader models. While recent Large Language Models (LLMs) like GPT-3 have demonstrated their effectiveness in zeroshot ODQA using direct prompting methods, these methods still fall short of fully harnessing the potential of LLMs when implicitly invoked. In this paper, we propose a Self- Prompting framework to explicitly utilize the massive knowledge encoded in the parameters of LLMs and their strong instruction understanding abilities. Concretely, we prompt LLMs step by step to generate multiple pseudo QA pairs with background passages and explanations entirely from scratch. These generated elements are then utilized for in-context learning. Experimental results show that our method significantly surpasses previous stateof- the-art zero-shot methods on three widelyused ODQA datasets and even achieves comparable performance with various customized fine-tuned models on full training data. Our code is available at https://***/ lockon-n/self-prompting. © 2022, CC BY-NC-SA.

关键词： Computational linguistics

Relational proofs for quantum programs

学校读者我要写书评

暂无评论

Relational proofs for quantum programs

Proceedings of the ACM on Programming Languages

作者： Barthe, Gilles Hsu, Justin Ying, Mingsheng Yu, Nengkun Zhou, Li Max Planck Institute for Security and Privacy Germany IMDEA Software Institute Spain Department of Computer Sciences University of Wisconsins Madison United States Centre for Quantum Software and Information University of Technology Sydney Australia State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Tsinghua University China Department of Computer Science and Technology Tsinghua University China

Relational verification of quantum programs has many potential applications in quantum and post-quantum security and other domains. We propose a relational program logic for quantum programs. The interpretation of our logic is based on a quantum analogue of probabilistic couplings. We use our logic to verify non-trivial relational properties of quantum programs, including uniformity for samples generated by the quantum Bernoulli factory, reliability of quantum teleportation against noise (bit and phase flip), security of quantum one-time pad and equivalence of quantum walks. © 2020 Copyright held by the owner/author(s).

关键词： Application programs computer circuits Couplings Verification Non trivial One time pads Post quantum securities Program logic Quantum programming Quantum teleportation Quantum walk Relational properties

Magnetization switching in polycrystalline Mn3Sn thin film induced by self-generated spin-polarized current

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Xie, Hang Chen, Xin Zhang, Qi Mu, Zhiqiang Zhang, Xinhai Yan, Binghai Wu, Yihong Department of Electrical and Computer Engineering National University of Singapore Singapore117583 Singapore Department of Electrical and Electronic Engineering Southern University of Science and Technology Xueyuan Rd. 1088 Shenzhen518055 China State Key Laboratory of Functional Materials for Informatics Shanghai Institute of Microsystem and Information Technology Chinese Academy of Sciences Shanghai200050 China Department of Condensed Matter Physics Weizmann Institute of Science Rehovot7610001 Israel

Electrical manipulation of spins is essential to design state-of-the-art spintronic devices and commonly relies on the spin current injected from a second heavy-metal material. The fact that chiral antiferromagnets produce spin current inspires us to explore the magnetization switching of chiral spins using self-generated spin torque. Here, we demonstrate the electric switching of noncollinear antiferromagnetic state in Mn3Sn by observing a crossover from conventional spin-orbit torque to the self-generated spin torque when increasing the MgO thickness in Ta/MgO/Mn3Sn polycrystalline films. The spin current injection from the Ta layer can be controlled and even blocked by varying the MgO thickness, but the switching sustains even at a large MgO thickness. Furthermore, the switching polarity reverses when the MgO thickness exceeds around 3 nm, which cannot be explained by the spin-orbit torque scenario due to spin current injection from the Ta layer. Evident current-induced switching is also observed in MgO/Mn3Sn and Ti/Mn3Sn bilayers, where external injection of spin Hall current to Mn3Sn is negligible. The inter-grain spin-transfer torque induced by spin-polarized current explains the experimental observations. Our findings provide an alternative pathway for electrical manipulation of non-collinear antiferromagnetic state without resorting to the conventional bilayer structure. © 2022, CC BY.

关键词： Magnesia