检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Zhao, Ganlong Li, Guanbin Pan, Jia Yu, Yizhou The University of Hong Kong Hong Kong Sun Yat-sen University China Guangdong Key Laboratory of Big Data Analysis and Processing China Peng Cheng Laboratory China

Aerial Vision-and-Language Navigation (Aerial VLN) aims to obtain an unmanned aerial vehicle agent to navigate aerial 3D environments following human instruction. Compared to ground-based VLN, aerial VLN requires the agent to decide the next action in both horizontal and vertical directions based on the first-person view observations. Previous methods struggle to perform well due to the longer navigation path, more complicated 3D scenes, and the neglect of the interplay between vertical and horizontal actions. In this paper, we propose a novel grid-based view selection framework that formulates aerial VLN action prediction as a grid-based view selection task, incorporating vertical action prediction in a manner that accounts for the coupling with horizontal actions, thereby enabling effective altitude adjustments. We further introduce a grid-based bird’s eye view map for aerial space to fuse the visual information in the navigation history, provide contextual scene information, and mitigate the impact of obstacles. Finally, a cross-modal transformer is adopted to explicitly align the long navigation history with the instruction. We demonstrate the superiority of our method in extensive experiments. Copyright © 2025, The Authors. All rights reserved.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous-attributes enhancement deep framework for network embedding

引用

Frontiers of Computer Science 2021年第6期15卷 121-131页

作者： Lisheng QIAO Fan ZHANG Xiaohui HUANG Kai LI Enhong CHEN Anhui Province Key Laboratory of Big Data Analysis and Application University of Science and Technology of ChinaHefei 230022China National Key Laboratory of Blind Signal Processing Chengdu 610041China School of Computer Science and Technology University of Science and Technology of ChinaHefei 230022China

Network embedding,which targets at learning the vector representation of vertices,has become a crucial issue in network ***,considering the complex structures and heterogeneous attributes in real-world networks,existing methods may fail to handle the inconsistencies between the structure topology and attribute ***,more comprehensive techniques are urgently required to capture the highly non-linear network structure and solve the existing inconsistencies with retaining more *** that end,in this paper,we propose a heterogeneous-attributes enhancement deep framework(HEDF),which could better capture the non-linear structure and associated information in a deep learningway,and effectively combine the structure information of multi-views by the combining *** this line,the inconsistencies will be handled to some extent and more structure information will be preserved through a semi-supervised *** extensive validations on several real-world datasets show that our model could outperform the baselines,especially for the sparse and inconsistent situation with less training data.

关键词： network embedding heterogeneous-attributes deep framework inconsistent

来源：评论

学校读者我要写书评

暂无评论

Tumor State-Space Network for High and Low Grade Glioma Classification 17

Tumor State-Space Network for High and Low Grade Glioma Clas...

引用

17th IEEE International Conference on Signal processing, ICSP 2024

作者： Chen, Qijian Wang, Lihui Deng, Zeyu Wang, Li Ye, Chen Zhu, YueMin Guizhou University Engineering Research Center of Text Computing Ministry of Education Key Laboratory of Intelligent Medical Image Analysis and Precise Diagnosis of Guizhou Province State Key Laboratory of Public Big Data College of Computer Science and Technology Guiyang China University Lyon Insa Lyon Cnrs Inserm Irp Metislab Creatis UMR5220 U1206 Lyon69621 France

ISBN: (纸本)9798350387384

Accurately predicting the grade of gliomas is crucial for choosing right treatment plans. While current methods using radiomics and deep learning can predict glioma grades effectively using magnetic resonance imaging (MRI), they overlook dynamic changes in the tumor and its neighboring areas. To address this issue, we propose a tumor state-space network (TSSNet), which fully exploits changes in the tumor region and its surroundings through dynamical updating strategy. The experimental results demonstrated that TSSNet achieves 90.32% accuracy and 93.55% area under the curve (AUC) for the prediction of high-grade and low-grade gliomas, respectively, which are 2.4% and 2.5% higher than the best state-of-the-art deep learning models. Moreover, the proposed state-space module can effectively enhance prediction performance, which is advantageous for personalized glioma treatment. © 2024 IEEE.

关键词： image enhancement

来源：评论

学校读者我要写书评

暂无评论

Design and implementation of algorithm execution time predictor for remote sensing product workflow 3

Design and implementation of algorithm execution time predic...

引用

3rd International Conference on Advanced Algorithms and Signal image processing, AASIP 2023

作者： Zhou, Bing Wu, Qiong Li, Jiaguo Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng475001 China School of Computer and Information Engineering Henan University Kaifeng475001 China Institute of Remote Sensing Applications Chinese Academy of Sciences Beijing100101 China

ISBN: (纸本)9781510668522

Remote sensing product production is an important task in the field of remote sensing engineering. Compared with the production process of remote sensing products, the implementation process of remote sensing product production is a multi-Task serial and highly concurrent process that integrates a large number of remote sensing digital image processing algorithms and various processes. Remote sensing product workflows transform complex computing tasks into processes that can be described, managed, and automated. In the workflow, algorithms are the smallest execution unit of tasks, and their execution time has a significant impact on the performance and efficiency of the entire workflow. Therefore, accurate prediction of algorithm execution time is crucial for optimizing workflow performance. This paper proposes and designs an algorithm execution time predictor for remote sensing product workflows. Firstly, feature extraction algorithms are used to extract features from algorithm execution logs, and the prediction model suitable for this scenario is determined by analyzing the characteristics of operators in remote sensing product workflows. Then, the model is trained using historical data sets, and the test set results are compared with the actual execution time to verify the accuracy and practicality of the model. Experimental results show that the proposed algorithm execution time predictor outperforms the baseline model in terms of prediction accuracy and time efficiency © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Remote sensing

来源：评论

学校读者我要写书评

暂无评论

Enhancing Event Tagger with Automatic Speech Recognizer for Audio Multi-task Scenarios by Distillation with Pre-Trained Large Models

Enhancing Event Tagger with Automatic Speech Recognizer for ...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Cheng, Jianfeng Liu, Ye Yin, Jian Wang, Liangdao Pan, Yan Lizhi Inc. Artificial Intelligence Department Guangzhou Beijing China School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China School of Artificial Intelligence Sun Yat-sen University Zhuhai China

ISBN: (纸本)9781665410205

With the continuous expansion of robotics and digital humans in practical applications, the demand for the auditory system is becoming deeper, usually requiring more efficient speech recognition framework capabilities to handle multiple tasks and use fewer resources. In previous audio processing frameworks, each type of audio processing task typically requires constructing a standalone deep network model for training, which results in more training data and higher training time when constructing models for multi-task audio scenarios simultaneously. The recent improvement of audio models based on transformers have brought about methods that can handle multiple audio tasks concurrently. However, recent related methods still require retraining multi-task targets with an amount of data, and achieve the general effect after training for multi-task scenarios than the simple combination of standalone methods processed separately. In order to better build a model that can handle multiple audio tasks, we propose a novel framework of distillation through pre-trained large models for enhancing event tagger with automatic speech recognizer. Through multiple rounds of experiments on several audio datasets, it has been verified that the proposed framework can achieve better results than the baseline for multitasking, and comparative results with less parameters compared to the baselines for single-task scenarios. © 2024 IEEE.

关键词： Continuous speech recognition

来源：评论

学校读者我要写书评

暂无评论

EEG-Based Fatigue Detection Using PLI Brain Network and Relief Algorithm

EEG-Based Fatigue Detection Using PLI Brain Network and Reli...

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： He, Yan Wang, Zhongmin Zhao, Yupeng Xi'an University of Posts and Telecommunications Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Xi'an Key Laboratory of Big Data and Intelligent Computing Shaanxi Xi'an710121 China School of Computer Science and Technology Xi'an University of Posts and Telecommunications Shaanxi Xi'an710121 China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

EEG-based fatigue driving monitoring has important application value in road traffic safety, and the ultimate goal of the research is the development and use of wearable devices, and too many EEG channels in practical application scenarios is detrimental to device portability, and it will lead to problems such as large amount of data, complex calculation and long processing time, so it is especially important to study how to select the EEG channels highly correlated with fatigue. In this paper, a PLI-Relief-based channel selection algorithm by combining the PLI functional connectivity and the weighting idea of Relief algorithm is proposed, and it is applied to the channel selection of fatigue driving EEG. First, the PLI functional connectivity matrix is constructed for the EEG signals after preprocessing, and the binarized PLI matrix is mapped into a brain functional network, and the prime channels are selected by the degree property of the brain network. Then, the power spectral density features are extracted from the EEG signals of the prime channels, and the weights of each prime channel are obtained using the relief algorithm, then the number and the names of optimal channels are determined according to the recognition accuracy of different channel combinations. The proposed method was validated on the publicly available SEED-VIG dataset, and the data of the seven optimal channels is finally selected and obtains a classification accuracy of 81.25%. The framework proposed in this paper takes into account both the correlation between channels and the characteristics of the channel signals themselves in channel selection, which is a reference value for the development and application of wearable devices. © 2022 IEEE.

关键词： Wearable technology

来源：评论

学校读者我要写书评

暂无评论

QoS Prediction based on the Low-rank Autoregressive Tensor Completion

QoS Prediction based on the Low-rank Autoregressive Tensor C...

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： Xia, Hong Dong, Qingyi Chen, Yanping Zheng, Jiahao Gao, Cong Wang, ZhongMin School of Computer Science and Technology Xi’an University of Posts and Telecommunications Shaanxi Xi’an710121 China Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Shaanxi Xi’an710121 China Xi'an Key Laboratory of Big Data and Intelligent Computing Shaanxi Xi’an710121 China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

With the rapid development of network services and edge computing, Quality of Service (QoS) has become an important indicator to validate performances of a network. Recommend high-quality services to users based on QoS values. However, the high sparsity of QoS data is because users usually call certain services only at a given time. Missing QoS data is very common in various service recommendation systems. Therefore, it is essential to predict QoS data to accurately recommend high-quality services to users. For the QoS data prediction problem, we build a third-order data tensor 'User-Service-Time' for the time series characteristics of QoS data. And introduce time-series variation as a regularization term into third-order tensor data prediction. We propose a QoS prediction framework using Low-Rank Autoregressive Tensor Completion (LATC). In particular, constructing a third-order tensor data model can better capture the global consistency of the data structure. Time regularization is introduced to take into account the local correlation of the data. Finally, in order to solve the constrained optimization problem, we use the general Alternating Direction Method of Multipliers (ADMM) to minimize the iterative optimization of variables and autoregressive parameters to obtain the final prediction result. Meanwhile, we conduct extensive research experiments on the real dataset WS-Dream. Experiments show that the QoS data prediction accuracy of our proposed QoS prediction method is higher than that of existing prediction methods under different degrees of data density. © 2022 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Entity disambiguation method based on Graph Attention Networks 14

Entity disambiguation method based on Graph Attention Networ...

引用

14th International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2022

作者： Qiang, Chengyu Li, Xiaogo Ma, Xianyan Han, Wei Yao, Yi School of Computer Science and Technology Xi'an University of Posts and Telecommunications Shaanxi Xi'an71012 China Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Shaanxi Xi'an710121 China Xi 'an Key Laboratory of Big Data and Intelligent Computing Shaanxi Xi'an710121 China

ISBN: (纸本)9781665499781

Entity disambiguation based on entity-link is a technique which constructs the mappings between entity reference items appearing in the short text and target entity in knowledge base respectively. In this paper we propose an entity disambiguation method based on Graph Attention Networks for semi-structured knowledge base data. First, a global Knowledge Graph is constructed from the semi-structured knowledge base, and the entity reference items are embedded by Bert pre-trained model meanwhile. Next, Graph Attention Networks which leverages masked self-attention layers is applyed on candidate entity nodes of global Knowledge Graph to fetch a vector of node level. Furtherly, we compute similarity scores rank between the entity reference items and the candidate entity to complete the task of entity disambiguation. The experimental results on CCKS2019 dataset achieve state-of-the-art. © 2022 IEEE

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

User Information Enhanced Knowledge Graph Convolutional Networks for Recommender Systems 14

User Information Enhanced Knowledge Graph Convolutional Netw...

引用

14th International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2022

作者： Hao, Junyu Li, Xiaoge School of Computer Science and Technology Xi'an University of Posts and Telecommunications Shaanxi Xi'an71012 China Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Shaanxi Xi'an710121 China Xi'an Key laboratory of Big Data and Intelligent Computing Shaanxi Xi'an710121 China

ISBN: (纸本)9781665499781

To deeply excavate the information contained in user data and better alleviate cold start problem, we propose KGCN++, a user information enhanced knowledge graph convolutional networks model for recommender system, which simulates nerves' influence diffusion of information of user and items in a unified framework. The model combines user information and user-item interaction information into a user knowledge graph, which advances KGCN by injecting high-order user potential information reflected in user graph and high-order influence of items reflected in item knowledge graph. In addition, we have designed the influence factors of different items on users, which are used to characterize the weight of items on the user's information dissemination, so as to distinguish users with different information. Through extensive experiments on real-world datasets, we demonstrate that KGCN++ achieves substantial gains in a variety of ecenarios, including movie and book recommendation, better performance compared to the other methods. © 2022 IEEE

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Mining Bytecode Features of Smart Contracts to Detect PonziScheme on Blockchain

引用

Computer Modeling in Engineering & Sciences 2021年第6期127卷 1069-1085页

作者： Xiajiong Shen Shuaimin Jiang Lei Zhang Henan Key Laboratory of Big Data Analysis and Processing Henan UniversityKaifeng47500China Institute of Data and Knowledge Engineering Henan UniversityKaifeng47500China School of Computer and Information Engineering Henan UniversityKaifeng47500China

The emergence of smart contracts has increased the attention of industry and academia to blockchain technology,which is tamper-proofing,decentralized,autonomous,and enables decentralized applications to operate in untrustworthy ***,these features of this technology are also easily exploited by unscrupulous individuals,a typical example of which is the Ponzi scheme in *** negative effect of unscrupulous individuals writing Ponzi scheme-type smart contracts in Ethereum and then using these contracts to scam large amounts of money has been *** solve this problem,we propose a detection model for detecting Ponzi schemes in smart contracts using *** this model,our innovation is shown in two aspects:We first propose to use two bytes as one characteristic,which can quickly transform the bytecode into a high-dimensional matrix,and this matrix contains all the implied characteristics in the ***,We innovatively transformed the Ponzi schemes detection into an anomaly detection ***,an anomaly detection algorithm is used to identify Ponzi schemes in smart *** results show that the proposed detection model can greatly improve the accuracy of the detection of the Ponzi scheme ***,the F1-score of this model can reach 0.88,which is far better than those of other traditional detection models.

关键词： Ponzi scheme blockchain security smart contracts anomaly detection

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：