检索结果-内蒙古大学图书馆

arXiv 2023年

作者： He, Yong Yu, Hongshan Yang, Zhengeng Liu, Xiaoyan Sun, Wei Mian, Ajmal National Engineering Laboratory for Robot Visual Perception and Control Technology College of Electrical and Information Engineering Hunan University Lushan South Rd. Yuelu Dist. Changsha410082 China Department of Computer Science The University of Western Australia WA6009 Australia

Point cloud processing methods exploit local point features and global context through aggregation which does not explicity model the internal correlations between local and global features. To address this problem, we propose full point encoding which is applicable to convolution and transformer architectures. Specifically, we propose Full Point Convolution (FPConv) and Full Point Transformer (FPTransformer) architectures. The key idea is to adaptively learn the weights from local and global geometric connections, where the connections are established through local and global correlation functions respectively. FPConv and FPTransformer simultaneously model the local and global geometric relationships as well as their internal correlations, demonstrating strong generalization ability and high performance. FPConv is incorporated in classical hierarchical network architectures to achieve local and global shape-aware learning. In FPTransformer, we introduce full point position encoding in self-attention, that hierarchically encodes each point position in the global and local receptive field. We also propose a shape aware downsampling block which takes into account the local shape and the global context. Experimental comparison to existing methods on benchmark datasets show the efficacy of FPConv and FPTransformer for semantic segmentation, object detection, classification, and normal estimation tasks. In particular, we achieve state-of-the-art semantic segmentation results of 76% mIoU on S3DIS 6-fold and 72.2% on S3DIS Area 5. Copyright © 2023, The Authors. All rights reserved.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

NeurAR: Neural Uncertainty for Autonomous 3D Reconstruction with Implicit Neural Representations

arXiv

引用

arXiv 2022年

作者： Ran, Yunlong Zeng, Jing He, Shibo Chen, Jiming Li, Lincheng Chen, Yingfeng Lee, Gimhee Ye, Qi The College of Control Science and Engineering Zhejiang University Hangzhou China The College of Control Science and Engineering The State Key Laboratory of Industrial Control Technology Zhejiang University The Key Laboratory of Collaborative Sensing and Autonomous Unmanned Systems of Zhejiang Province China Fuxi AI Lab NetEase Inc. Hangzhou China The Department of Computer Science National University of Singapore Singapore

Implicit neural representations have shown compelling results in offline 3D reconstruction and also recently demonstrated the potential for online SLAM systems. However, applying them to autonomous 3D reconstruction, where a robot is required to explore a scene and plan a view path for the reconstruction, has not been studied. In this paper, we explore for the first time the possibility of using implicit neural representations for autonomous 3D scene reconstruction by addressing two key challenges: 1) seeking a criterion to measure the quality of the candidate viewpoints for the view planning based on the new representations, and 2) learning the criterion from data that can generalize to different scenes instead of a hand-crafting one. To solve the challenges, firstly, a proxy of Peak Signal-to-Noise Ratio (PSNR) is proposed to quantify a viewpoint quality;secondly, the proxy is optimized jointly with the parameters of an implicit neural network for the scene. With the proposed view quality criterion from neural networks (termed as Neural Uncertainty), we can then apply implicit representations to autonomous 3D reconstruction. Our method demonstrates significant improvements on various metrics for the rendered image quality and the geometry quality of the reconstructed 3D models when compared with variants using TSDF or reconstruction without view planning. Project webpage https://***/NeurAR/. Copyright © 2022, The Authors. All rights reserved.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Multi-tensor network representation for high-order tensor completion

arXiv

引用

arXiv 2021年

作者： Nie, Chang Wang, Huan Lai, Zhihui Nanjing210014 China College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China

—This work studies the problem of high-dimensional data (referred to tensors) completion from partially observed samplings. We consider that a tensor is a superposition of multiple low-rank components. In particular, each component can be represented as multilinear connections over several latent factors and naturally mapped to a specific tensor network (TN) topology. In this paper, we propose a fundamental tensor decomposition (TD) framework: Multi-Tensor Network Representation (MTNR), which can be regarded as a linear combination of a range of TD models, e.g., CANDECOMP/PARAFAC (CP) decomposition, Tensor Train (TT), and Tensor Ring (TR). Specifically, MTNR represents a high-order tensor as the addition of multiple TN models, and the topology of each TN is automatically generated instead of manually pre-designed. For the optimization phase, an adaptive topology learning (ATL) algorithm is presented to obtain latent factors of each TN based on a rank incremental strategy and a projection error measurement strategy. In addition, we theoretically establish the fundamental multilinear operations for the tensors with TN representation, and reveal the structural transformation of MTNR to a single TN. Finally, MTNR is applied to a typical task, tensor completion, and two effective algorithms are proposed for the exact recovery of incomplete data based on the Alternating Least Squares (ALS) scheme and Alternating Direction Method of Multiplier (ADMM) framework. Extensive numerical experiments on synthetic data and real-world datasets demonstrate the effectiveness of MTNR compared with the start-of-the-art methods. Copyright © 2021, 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

Machine learning assisted signal detection in ambient backscatter communication networks

TechRxiv

引用

TechRxiv 2021年

作者： Toro, Usman Saleh ElHalawany, Basem M. Wong, Aslan B. Wang, Lu Wu, Kaishun College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China Benha University Egypt

Ambient backscatter communication (AmBC) has emerged as a promising paradigm for enabling sustainable low-power operation of Internet of Things (IoT) devices. This is due to its ability to enable sensing and communication through backscattering ambient wireless signals (such as WiFi and TV signals). But a great impediment to AmBC-enabled networks is the difficulty in decoding the backscatter signals because the ambient signals are usually modulated and meant for other legacy receivers rather than AmBC devices. Drawing from the ability of machine learning (ML) to enhance the performance of wireless communication systems, some ML-aided techniques have been developed to assist signal detection in AmBC. Hence, this paper aims to provide a comprehensive overview of the subject by;describing the operation of the AmBC network, highlighting the major challenges to signal detection in AmBC, discussing and comparing the performance of some existing ML-assisted solutions to AmBC signal detection, and highlighting some future research that could be carried out on the subject. © 2021, CC BY.

关键词： Signal detection

来源：评论

学校读者我要写书评

暂无评论

BACON: Boundary-guided Polypharmacy Side Effect Prediction via Integrating Molecular Structures and Biochemical Information

BACON: Boundary-guided Polypharmacy Side Effect Prediction v...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Yike Wang Huifang Ma Zihao Gao Zhixin Li Liang Chang College of Computer Science and Engineering Northwest Normal University Lanzhou China Key Lab of Education Blockchain and Intelligent Technology Ministry of Education Guangxi Normal University Guilin Guangxi Key Laboratory of Trusted Software Guilin University of Electronic Technology Guilin

Polypharmacy side effect prediction is a vital task in healthcare machine learning, aiming to predict the occurrence of multiple side effects when patients take drugs together. Existing researches mainly follow the deep encoder-decoder paradigm and suffer from the following two limitations: 1) The encoder simply combines inadequate information about drugs; 2) The decoder fails to capture dependencies among different side effects, thus hindering accurate prediction. To overcome the aforementioned limitations, we propose a boundary-guided polypharmacy side effect prediction method (BACON). Our framework constructs two complementary views based on drug’s chemical substructure and biochemical features and enhances drug representations via contrastive learning. The decoder incorporates a boundary-guided strategy to capture drug interaction dependencies for optimizing polypharmacy side effect prediction. Experimental results demonstrate BACON superiority over SOTA models in accurately predicting drug side effect events.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Large Deviation Algorithm for the Thresholding Bandit Problem

SSRN

引用

SSRN 2023年

作者： He, Yulin Zhang, Manjing Liu, Guangwu Fournier-Viger, Philippe Dai, Shan Huang, Joshua Zhexue Shenzhen518107 China Department of Management Sciences City University of Hong Kong 518057 Hong Kong College of Computer Science & Software Engineering Shenzhen University Shenzhen518060 China Shenzhen Research Institute of Big Data Shenzhen518172 China

The thresholding bandit (TB) problem is a popular sequential decision-making problem, which aims at identifying the systems whose means are greater than a threshold. The TB problem is a variant of the multi-armed bandit (MAB) problem, which seeks to maximize the total reward by balancing the trade-off between exploration and exploitation. This paper presents a parameter-free policy for the TB problem based on the large-deviation theory, and proposes a large-deviation (LD) algorithm to solve the TB problem using an optimal allocation rule. We theoretically show that the LD algorithm is asymptotically optimal, and its probability of incorrect identification converges more quickly than the existing algorithms as the sample size approaches to infinity. Furthermore, to improve the robustness of the LD algorithm for optimization problems with small sample sizes, a decay rate-based approximate large-deviation (ALD) algorithm is proposed that takes the sample means and sample variances into account in a direct and simple way. Extensive experiments are conducted to validate the effectiveness of the LD and ALD algorithms. Simulation results indicate that both LD and ALD are viable solutions to handle the TB problem. © 2023, The Authors. All rights reserved.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Multi-block-single-probe variance reduced estimator for coupled compositional optimization 22

Multi-block-single-probe variance reduced estimator for coup...

引用

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Wei Jiang Gang Li Yibo Wang Lijun Zhang Tianbao Yang National Key Laboratory for Novel Software Technology Nanjing University Nanjing China Department of Computer Science the University of Iowa Iowa City Department of Computer Science and Engineering Texas A&M University College Station

ISBN: (纸本)9781713871088

Variance reduction techniques such as SPIDER/SARAH/STORM have been extensively studied to improve the convergence rates of stochastic non-convex optimization, which usually maintain and update a sequence of estimators for a single function across iterations. What if we need to track multiple functional mappings across iterations but only with access to stochastic samples of O(1) functional mappings at each iteration? There is an important application in solving an emerging family of coupled compositional optimization problems in the form of ∑mi=1 fi(gi(w)), where gi is accessible through a stochastic oracle. The key issue is to track and estimate a sequence of g(w) = (g1(w),..., gm(w)) across iterations, where g(w) has m blocks and it is only allowed to probe O(1) blocks to attain their stochastic values and Jacobians. To improve the complexity for solving these problems, we propose a novel stochastic method named Multi-block-Single-probe Variance Reduced (MSVR) estimator to track the sequence of g(w). It is inspired by STORM but introduces a customized error correction term to alleviate the noise not only in stochastic samples for the selected blocks but also in those blocks that are not sampled. With the help of the MSVR estimator, we develop several algorithms for solving the aforementioned compositional problems with improved complexities across a spectrum of settings with non-convex/convex/strongly convex/Polyak-Łojasiewicz (PL) objectives. Our results improve upon prior ones in several aspects, including the order of sample complexities and dependence on the strong convexity parameter. Empirical studies on multi-task deep AUC maximization demonstrate the better performance of using the new estimator.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Image de-blurring Model based on Machine Learning

Image de-blurring Model based on Machine Learning

引用

2021 Asia-Pacific Conference on Image Processing, Electronics and computers, IPEC 2021

作者： Li, Junsheng Xu, Silu Wang, Cong College of Computer Science and Technology Harbin Engineering University Heilongjiang Harbin150001 China City Institute Dalian University of Technology Liaoning Dalian1166000 China School of Information and Control Engineering Qingdao University of Technology Shandong Qingdao266520 China

Due to the influence of haze in winter, outdoor images usually lose contrast and fidelity. In view of the fact that most de-fog algorithms are not effective for images with large sky areas, an improved dark channel a priori de-fog method is proposed. First of all, the sky region is segmented according to the image gradient information, and on the basis of sky segmentation, the atmospheric light value is reasonably estimated by setting the discriminant formula combined with the high brightness and smoothness of atmospheric light reference pixels. Secondly, according to the different dark channel values, the piecewise linear function is used to dynamically modify the adjustable parameters to solve the local shadow caused by excessive defog. Then, the transmittance estimated by the bright channel model and the improved dark channel prior model are fused, and the edge is optimized by guided filtering. Finally, combined with the atmospheric scattering model, the defog image is obtained by brightness compensation and contrast stretching. The experimental results show that the improved method can effectively improve the image distortion, enhance the image contrast and details, especially in maintaining the visual authenticity of the sky region. © Published under licence by IOP Publishing Ltd.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

A Survey of Data Partitioning and Sampling Methods to Support Big Data Analysis

引用

Big Data Mining and Analytics 2020年第2期3卷 85-101页

作者： Mohammad Sultan Mahmud Joshua Zhexue Huang Salman Salloum Tamer Z.Emara Kuanishbay Sadatdiynov National Engineering Laboratory for Big Data System Computing Technology Shenzhen UniversityUniversityShenzhen 518060China Big Data Institute College of Computer Science and Software EngineeringShenzhen UniversityShenzhen 518060China.

computer clusters with the shared-nothing architecture are the major computing platforms for big data processing and *** cluster computing,data partitioning and sampling are two fundamental strategies to speed up the computation of big data and increase *** this paper,we present a comprehensive survey of the methods and techniques of data partitioning and sampling with respect to big data processing and *** start with an overview of the mainstream big data frameworks on Hadoop *** basic methods of data partitioning are then discussed including three classical horizontal partitioning schemes:range,hash,and random *** partitioning on Hadoop clusters is also discussed with a summary of new strategies for big data partitioning,including the new Random Sample Partition(RSP)distributed *** classical methods of data sampling are then investigated,including simple random sampling,stratified sampling,and reservoir *** common methods of big data sampling on computing clusters are also discussed:record-level sampling and blocklevel ***-level sampling is not as efficient as block-level sampling on big distributed *** the other hand,block-level sampling on data blocks generated with the classical data partitioning methods does not necessarily produce good representative samples for approximate computing of big *** this survey,we also summarize the prevailing strategies and related work on sampling-based approximation on Hadoop *** believe that data partitioning and sampling should be considered together to build approximate cluster computing frameworks that are reliable in both the computational and statistical respects.

关键词： big data analysis data partitioning data sampling distributed and parallel computing approximate computing

来源：评论

学校读者我要写书评

暂无评论

Boosting component-based synthesis with control structure recommendation 1

Boosting component-based synthesis with control structure re...

引用

1st ACM SIGSOFT International Workshop on Representation Learning for software engineering and Program Languages, RL+SE and PL 2020, co-located with ESEC/FSE 2020

作者： Liu, Binbin Dong, Wei Zhang, Yating Wang, Daiyan Liu, Jiaxin College of Computer Science National University of Defense Technology Changsha China Key Lab. of Software Engineering for Complex Systems College of Computer Science National University of Defense Technology Changsha China

ISBN: (纸本)9781450381253

Component-based synthesis is an important research field in program synthesis. API-based synthesis is a subfield of component-based synthesis, the component library of which are Java APIs. Unlike existing work in API-based synthesis that can only generate loop-free programs constituted by APIs, state-of-The-Art work FrAngel can generate programs with control structures. However, for the generation of control structures, it samples different types of control structures all at random. Given the information about the desired method (such as method name and input/output types), experienced programmers can have an initial thought about the possible control structures that could be used in implementing the desired method. The knowledge about control structures in the method can be learned from high-quality projects. In this paper, we propose a novel approach of recommending control structures for API-based synthesis based on deep learning. A neural network that can jointly embed the natural language description, method name, and input/output types into high-dimensional vectors to predict the possible control structures of the desired method is proposed. We integrate the prediction model into the synthesizer to improve the efficiency of synthesis. We train our model on a codebase of high-quality Java projects from GitHub. The prediction results of the neural network are fed to the API-based synthesizer to guide the sampling process of control structures. The experimental results on 40 programming tasks show that our approach can effectively improve the efficiency of synthesis. © 2020 ACM.

关键词： Efficiency

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

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

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

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

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：