检索结果-内蒙古大学图书馆

IEEE 24th International Conference on High Performance Switching and Routing, (HPSR)

作者： Peng, Yufei Guo, Yingya Hao, Run Lin, Junda Fuzhou Univ Dept Comp & Data Sci Fuzhou Peoples R China Fuzhou Univ Fujian Prov Key Lab Network Comp & Intelligent In Fuzhou Peoples R China

ISBN: (纸本)9781665476409

Network traffic prediction plays a significant role in network management. Previous network traffic prediction methods mainly focus on the temporal relationship between network traffic, and used time series models to predict network traffic, ignoring the spatial information contained in traffic data. Therefore, the prediction accuracy is limited, especially in long-term prediction. To improve the prediction accuracy of the dynamic network traffic in the long term, we propose an Attention-based Spatial-Temporal Graph Network (ASTGN) model for network traffic prediction to better capture both the temporal and spatial relations between the network traffic. Specifically, in ASTGN, we exploit an encoder-decoder architecture, where the encoder encodes the input network traffic and the decoder outputs the predicted network traffic sequences, integrating the temporal and spatial information of the network traffic data through the Spatio-Temporal Embedding module. The experimental results demonstrate the superiority of our proposed method ASTGN in long-term prediction.

关键词： Network Traffic Prediction Attention Mechanism Graph Neural Network Temporal and Spatial encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Multi-Service Demand Forecasting using Graph Neural Networks 17

Multi-Service Demand Forecasting using Graph Neural Networks

引用

17th IEEE International Congress on Intelligent and Service-Oriented Systems Engineering (CISOSE)

作者： Theodoropoulos, Theodoros Makris, Antonios Kontopoulos, Ioannis Maroudis, Angelos-Christos Tserpes, Konstantinos Harokopio Univ Athens Dept Informat & Telemat Kallithea Greece

ISBN: (纸本)9798350322392

Accurate service demand forecasting is crucial towards achieving effective resource allocation and service orchestration. However, existing solutions require separate prediction models for each service, which consumes significant computation resources. In this paper, we propose a novel approach that introduces a global prediction model capable of generating accurate multi-step predictions involving multiple services and that can potentially leveraged for facilitating proactive strategies for orchestrating multiple services. The proposed model is based on an encoder-decoder architecture that utilizes Graph Neural Networks (GNNs) to capture interdependencies among input variables and their trends. By iteratively updating graph node representations, our model effectively incorporates historical trends and potential dependencies. Furthermore, the incorporation of GNNs into the encoder-decoder architecture enables the proposed approach to leverage the correlations among input variables, thus making it suitable for multivariate forecasting. Experimental results demonstrate the superiority of the proposed approach in terms of accurately conducting multi-step multiservice demand predictions, when compared against numerous contemporary deep learning models.

关键词： graph neural networks service demand forecasting edge deep learning encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

BDMFuse: Multi-scale network fusion for infrared and visible images based on base and detail features

引用

JOURNAL OF INFRARED AND MILLIMETER WAVES 2025年第2期44卷 275-284页

作者： Si, Hai-Ping Zhao, Wen-Rui Li, Ting-Ting Li, Fei-Tao Fernado, Bacao Sun, Chang-Xia Li, Yan-Ling Henan Agr Univ Coll Informat & Management Sci Zhengzhou 450046 Peoples R China Univ Nova Lisboa NOVA Informat Management Sch P-1070312 Lisbon Portugal

The fusion of infrared and visible images should emphasize the salient targets in the infrared image while preserving the textural details of the visible images. To meet these requirements, an autoencoder-based method for infrared and visible image fusion is proposed. The encoder designed according to the optimization objective consists of a base encoder and a detail encoder, which is used to extract low-frequency and high-frequency information from the image. This extraction may lead to some information not being captured, so a compensation encoder is proposed to supplement the missing information. Multi-scale decomposition is also employed to extract image features more comprehensively. The decoder combines low-frequency, high-frequency and supplementary information to obtain multi-scale features. Subsequently, the attention strategy and fusion module are introduced to perform multi-scale fusion for image reconstruction. Experimental results on three datasets show that the fused images generated by this network effectively retain salient targets while being more consistent with human visual perception.

关键词： infrared image visible image image fusion encoder-decoder multi-scale features

来源：评论

学校读者我要写书评

暂无评论

RBBA: ResNet - BERT - Bahdanau Attention for Image Caption Generator 14

RBBA: ResNet - BERT - Bahdanau Attention for Image Caption G...

引用

14th International Conference on Information and Communication Technology Convergence, ICTC 2023

作者： Hoang, Duc-Hieu Tran, Anh-Khoa Minh Dang, Duc Ngoc Tran, Phuong-Nam Dang-Ngoc, Hanh Nguyen, Cuong Tuan Ton Duc Thang University Faculty of Electrical & Electronics Eng. Ho Chi Minh City Viet Nam Ton Duc Thang University Faculty of Electrical and Electronics Ho Chi Minh City Viet Nam Fpt University Dept. of Computing Fundamental Ho Chi Minh City Viet Nam University of Technology Sydney School of Electrical and Data Engineering SydneyNSW Australia Fpt University Dept. of Information Technology Specialization Ho Chi Minh City Viet Nam

ISBN: (纸本)9798350313277

In recent years, the topic of image caption generators has gained significant attention. Several successful projects have emerged in this field, showcasing notable advancements. Image caption generators automatically generate descriptive captions for images through the encoder and decoder mechanisms. The encoder leverages computer vision models, while the decoder utilizes natural language processing models. In this study, we aim to assess a comprehensive set of seven distinct methodologies, including six existing methods from prior research and one newly proposed. These methods are trained and evaluated with bilingual evaluation (BLEU) on the Flickr8K dataset. In our experiments, the proposed ResNet50 - BERT - Bahdanau Attention model outperforms the other models in terms of the BLEU-1 score of 0.532143 and BLEU-4 score of 0.126316. © 2023 IEEE.

关键词： BLEU Deep learning encoder-decoder Flickr8K Image Caption Natural language processing

来源：评论

学校读者我要写书评

暂无评论

Improving Long-Term Traffic Prediction with Online Search Log Data 47

Improving Long-Term Traffic Prediction with Online Search Lo...

引用

47th IEEE-Computer-Society Annual International Conference on Computers, Software, and Applications (COMPSAC)

作者： Matsunaga, Itsuki Kosugi, Yuto Ge Hangli Michikata, Takashi Koshizuka, Noboru Univ Tokyo Grad Sch Interdisciplinary Informat Studies Tokyo Japan Univ Tokyo Interfac Initiat Informat Studies Tokyo Japan

ISBN: (纸本)9798350326970

Long-term traffic prediction is essential for both road managers and users to prepare for future congestion. However, most existing studies have only focused on short-term prediction. Moreover, few studies have effectively incorporated external data into long-term traffic prediction, even though traffic conditions are complexly influenced by various spatiotemporal factors. In this paper, we propose a novel method that utilizes online search log data for long-term traffic prediction on expressways. Online search logs reflect drivers' travel intentions and external factors, such as weather conditions and events, which cannot be represented by historical traffic data. Based on a new analysis of the correlation between online search log data and real-world traffic, we use online search log data as potential future traffic volume in an LSTM-based encoder-decoder model. Experiments using a real-world dataset on an expressway known for frequent congestion show that the use of online search log data improves the metrics of MAE, RMSE, and MAPE in next-day traffic volume prediction by 8.1%, 12.5%, and 7.2% on average, respectively. Similarly, in speed prediction, the MAE, RMSE, and MAPE are reduced by 3.7%, 2.1%, and 11.8%, respectively. It is also shown that online search log data is particularly effective in predicting irregular congestion caused by sudden increases in traffic demand.

关键词： Traffic prediction online search log intelligent transportation system encoder-decoder spatiotemporal data

来源：评论

学校读者我要写书评

暂无评论

An encoder framework for taxi-demand prediction using spatio-temporal function approximation

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2023年第PartA期126卷

作者： Bhanu, Manish Roy, Saswata Priya, Shalini Mendes-Moreira, Joao Chandra, Joydeep Indian Inst Technol Patna Dept Comp Sci & Engn Dayalpur Daulatpur Bihar India Oak Ridge Natl Lab Oak Ridge TN USA Univ Porto Fac Engn Dept Informat Engn Porto Portugal

Predicting taxi demands in large cities can help in better traffic management as well as ensure better commuter satisfaction for an intelligent transportation system. However, the traffic demands across different locations have varying spatio-temporal correlations that are difficult to model. Despite the ability of the existing Deep Neural Network (DNN) models to capture the non-linearity in spatial and temporal characteristics of the demand time-series, capturing spatio-temporal characteristics in different real-world scenarios like varying historic and prediction time frame, spatio-temporal variations due to noise or missing data, etc. still remain a big challenge for the state-of-the-art models. In this paper, we introduce encoder-ApproXimator (EnAppX), an encoder-decoder DNN-based model that uses Chebyshev function approximation in the decoding stage for taxi demand times-series prediction and can better estimate the time-series in the presence of large spatio-temporal variations. Opposed to any existing state-of-the-art model, the proposed model approximates complete spatiotemporal characteristics in the frequency domain which in turn enables the model to make a robust and improved prediction in different scenarios. Validation over two real-world taxi datasets from different cities shows a considerable improvement of around 23% in RMSE scores compared to the state-of-the-art baseline model. Unlike several existing state-of-the-art models, EnAppX also produces improved prediction accuracy across two regions for both to and fro demands.

关键词： Spatio-temporal encoder-decoder Multi-step prediction Time-series Taxi-demand

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Code-switched Text Generation from Parallel Text 24

Unsupervised Code-switched Text Generation from Parallel Tex...

引用

Interspeech Conference

作者： Chi, Jie Lu, Brian Eisner, Jason Bell, Peter Jyothi, Preethi Ali, Ahmed M. Univ Edinburgh Ctr Speech Technol Res Edinburgh Midlothian Scotland Johns Hopkins Univ Ctr Language & Speech Proc Baltimore MD USA Indian Inst Technol Dept Comp Sci Bombay Maharashtra India HBKU Qatar Comp Res Inst Ar Rayyan Qatar

There has been great interest in developing automatic speech recognition (ASR) systems that can handle code-switched (CS) speech to meet the needs of a growing bilingual population. However, existing datasets are limited in size. It is expensive and difficult to collect real transcribed spoken CS data due to the challenges of finding and identifying CS data in the wild. As a result, many attempts have been made to generate synthetic CS data. Existing methods either require the existence of CS data during training, or are driven by linguistic knowledge. We introduce a novel approach of forcing a multilingual MT system that was trained on non-CS data to generate CS translations. Comparing against two prior methods, we show that simply leveraging the shared representations of two languages (Mandarin and English) yields better CS text generation and, ultimately, better CS ASR.

关键词： code-switching text generation data augmentation encoder-decoder unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

A Lightweight and Dynamic Convolutional Network for Real-time Semantic Segmentation 35

A Lightweight and Dynamic Convolutional Network for Real-tim...

引用

35th Chinese Control and Decision Conference (CCDC)

作者： Zhang, Chunyu Xu, Fang Wu, Chengdong Northeastern Univ Fac Robot Sci & Engn Shenyang Peoples R China Shenyang Siasun Robot & Automat Co Ltd Shenyang Peoples R China

ISBN: (纸本)9798350334722

Semantic segmentation is a difficult task that satisfies most of the demands of autonomous driving and drone aerial photography in a unified fashion. Convolutional neural networks can properly categorize image pixels via end-to-end model training. However, achieving the optimal trade-off between segmentation precision and the number of network parameters while maintaining a suitable inference time has become a challenging task. In this paper, we propose a lightweight dynamic convolutional semantic segmentation network, LDCNet, which belongs to the asymmetric network architecture. First, we designed a coding module that includes dynamic convolution: DDAB. The success of this module is attributed to the use of dynamic convolution, which increases the utilization of local and contextual information of the features. We also designed the decoding module containing feature pyramids and hybrid attention: HA-FP,which performs a multi-scale fusion of features accompanied by feature selection. On the Cityscapes and Camvid datasets, LDCNet obtains 73.5 mIoU and 69.4 mIoU accuracy with 78.4 FPS and 91.3 FPS, respectively, without pre-training or post-processing. Our experimental findings reveal that LDCNet achieves an outstanding balance between segmentation accuracy and network parameters with just 0.96 M.

关键词： Lightweight network Dynamic convolution encoder-decoder Semantic segmentation Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Enriching Grammatical Error Correction Resources for Modern Greek 13

Enriching Grammatical Error Correction Resources for Modern ...

引用

13th International Conference on Language Resources and Evaluation (LREC)

作者： Korre, Katerina Pavlopoulos, John Univ Bologna Forli Italy Athens Univ Econ & Business Athens Greece

ISBN: (纸本)9791095546726

Grammatical Error Correction (GEC), a task of Natural Language Processing (NLP), is challenging for under-represented languages. This issue is most prominent in languages other than English. This paper addresses the issue of data and system sparsity for GEC purposes in the modern Greek Language. Following the most popular current approaches in GEC, we develop and test an MT5 multilingual text-to-text transformer for Greek. To our knowledge this the first attempt to create a fully-fledged GEC model for Greek. Our evaluation shows that our system reaches up to 52.63% F0.5 score on part of the Greek Native Corpus (GNC), which is 16% below the winning system of the BEA-19 shared task on English GEC. In addition, we provide an extended version of the Greek Learner Corpus (GLC), on which our model reaches up to 22.76% F0.5. Previous versions did not include corrections with the annotations which hindered the potential development of efficient GEC systems. For that reason we provide a new set of corrections. This new dataset facilitates an exploration of the generalisation abilities and robustness of our system, given that the assessment is conducted on learner data while the training on native data.

关键词： GEC MT5 encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Crop classification for UAV visible imagery using deep semantic segmentation methods

引用

GEOCARTO INTERNATIONAL 2022年第25期37卷 10033-10057页

作者： Zhang, Shiqi Dai, Xiaoai Li, Jingzhong Gao, Xiaojie Zhang, Fuxi Gong, Fanxi Lu, Heng Wang, Meilian Ji, Fujiang Wang, Zekun Peng, Peihao Chengdu Univ Technol Coll Earth Sci Chengdu Peoples R China Wuhan Univ Sch Resource & Environm Sci Wuhan Peoples R China North Carolina State Univ Ctr Geospatial Analyt Raleigh NC USA Shanghai Ocean Univ Coll Engn Sci & Technol Shanghai Peoples R China Sichuan Univ State Key Lab Hydraul & Mt River Engn Chengdu Peoples R China Sichuan Univ Coll Hydraul & Hydroelect Engn Chengdu Peoples R China Hong Kong Polytech Univ Hong Kong Peoples R China Chinese Acad Sci Aerosp Informat Res Inst Beijing Peoples R China Univ Chinese Acad Sci Beijing Peoples R China Auburn Univ Dept Mech Engn Auburn AL 36849 USA Chengdu Univ Technol Tourism & Urban Rural Planning Coll Chengdu Peoples R China

Unmanned aerial vehicle (UAV) has become a mainstream data collection platform in precision agriculture. For more accessible UAV-visible imagery, the high spatial resolution brings the rich geometric texture features triggered large differences in same crop image's features. We proposed an encoder-decoder's fully convolutional neural network combined with a visible band difference vegetation index (VDVI) to perform deep semantic segmentation of crop image features. This model ensures the accuracy and the generalization ability, while reducing parameters and the operation cost. A case study of crop classification was conducted in Chengdu, China, where classified four crops, namely, maize, rice, balsam pear, and Loropetalum chinese, it was shown more effective results. In addition, this study explores a fine crop classification method based on visible light features, which is feasible with low equipment cost, and has a prospect of application in crop survey based on UAV low altitude remote sensing.

关键词： Crop UAV visible images deep semantic segmentation encoder-decoder

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：