检索结果-内蒙古大学图书馆

Evaluation of map-matching algorithms for smartphone-based active travel data

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第1期17卷 227-242页

作者： Berjisian, Elmira Bigazzi, Alexander Univ British Columbia Dept Civil Engn 6250 Appl Sci Lane Vancouver BC V6T 1Z4 Canada

Global Positioning System (GPS) data on walking and cycling trips can generate useful insights for transportation systems but require substantial processing. One of the key GPS data processing steps is "map-matching", or inference of the sequence of network links traversed during travel. The objective of this research is to evaluate the accuracy of existing map-matching algorithms for GPS data on active travel. A method to flag erroneous map-matching results without requiring ground-truth data and improvements for active travel data are also proposed. Six map-matching algorithms are applied to a sample of 63 trajectories, stratified on network density and average heading change, extracted from a large set of real-world trips from metropolitan Vancouver, Canada. Results show that the best performing method is PgMapMatch, which can be further improved by adjustments to link costs and allowing wrong-way travel. Two other algorithms have similarly accurate routes (70-90% accuracy, depending on the measure), but fail to generate routes for about a third of trips. The proposed error detection measure can be used (without ground truth data) to flag matched routes requiring visual inspection, with a recommendation to look for: Wrong-way travel, missing links in the network data, and parallel facilities on the same street.

关键词： network data transportation Mobile, ubiquitous and pervasive computing cycling trips substantial processing transportation systems geographic information systems key GPS data processing steps Geography and cartography computing walking flag matched routes smartphone-based active travel data flag erroneous map-matching results cartography traffic engineering computing smart phones Radionavigation and direction finding Global Positioning System network links existing map-matching algorithms ground-truth data ground truth data

来源：评论

学校读者我要写书评

暂无评论

Pedestrians crossing intention anticipation based on dual-channel action recognition and hierarchical environmental context

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第2期17卷 255-269页

作者： Ni, Rongrong Yang, Biao Wei, Zhiwen Hu, Hongyu Yang, Changchun Jilin Univ State Key Lab Automot Simulat & Control Renmin St 5988 Changchun Peoples R China Hohai Univ Coll IoT Engn Changzhou Peoples R China Changzhou Univ Sch Microelect & Control Engn Changzhou Peoples R China

The increase in car ownership has brought convenience to people's lives, but it has also led to a sharp increase in pedestrian-vehicle conflicts. Such conflicts can be avoided by braking when pedestrians in front of the car are detected. However, frequent braking reduces driving comfort and travel efficiency. Considering that pedestrian-vehicle conflicts always occur when pedestrians cross the road, this study proposes a multi-factor fusion network (MFFN) to anticipate pedestrians' crossing intentions. MFFN contains a dual-channel action recognition sub-network to robustly identify pedestrians' actions by adaptively fusing pedestrians' skeleton and representation characteristics. Afterward, object-level and semantic-level perception of traffic scenes are realised by a hierarchical graph attention network and a lightweight semantic segmentation method, respectively. These hierarchical environmental contexts can provide critical clues for identifying pedestrians' crossing intentions. Then, a self-attention mechanism is used to integrate various factors and produce intention identification results. The proposed MFFN is evaluated on public datasets PIE and JAAD. Experimental results show that the proposed method can accurately and reliably identify pedestrian crossing intentions. The accuracy, F1 score, AUC, and precision indicators on JAAD/PIE datasets are 0.912/0.876, 0.813/0.806, 0.896/0.889, and 0.802/0.788, respectively.

关键词： traffic engineering computing image segmentation feature extraction recurrent neural nets hierarchical environmental context road safety learning (artificial intelligence) object-level Neural nets intention anticipation Optical, image and video signal processing Computer vision and image processing techniques dual-channel action recognition convolutional neural nets pedestrian-vehicle conflicts traffic engineering computing object detection multifactor fusion network deep learning (artificial intelligence) pedestrians Image recognition graph theory image sequences road traffic semantic-level perception MFFN pedestrian crossing intentions road vehicles image motion analysis Combinatorial mathematics hierarchical graph attention network

来源：评论

学校读者我要写书评

暂无评论

Short-term traffic flow prediction approach incorporating vehicle functions from RFID-ELP data for urban road sections

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第1期17卷 144-164页

作者： Huang, Shuai Sun, Dihua Zhao, Min Chen, Jin Chen, Rui Chongqing Univ Key Lab Dependable Serv Comp Cyber Phys Soc Minist Educ Chongqing 400044 Peoples R China Chongqing Univ Sch Automat Chongqing Peoples R China

Short-term traffic flow prediction is particularly important for urban traffic control and congestion management. To improve the accuracy of traffic flow prediction, a new traffic flow prediction approach is proposed from the perspective of vehicle functions by using Radio Frequency Identification (RFID) electronic license plate (ELP) data. First, the improved Wavelet Neural Network (WNN) is introduced by adding three momentum terms to avoid local minimum and overfitting. The Long Short-Term Memory Neural Network (LSTM) is constructed to train the time series with long time lags. Second, a novel short-term traffic flow prediction method is developed by combining the improved WNN and LSTM in an urban road section. The proposed method performs a weighted combination to obtain a new prediction value according to whether the difference between the prediction values of the two methods satisfy the defined threshold, and then selects the one with the smallest error among the three prediction values as the final prediction result. Third, the experiments explore the prediction performance of the weekdays and weekends with 5 min-interval, 15 min-interval and 30 min-interval via using RFID-ELP data from Chongqing, China. Compared with existing prediction methods, the proposed approach can promote prediction performance and meet real-time requirements.

关键词： radiofrequency identification traffic engineering computing prediction performance long time lags recurrent neural nets traffic control Time series urban road section time series Neural nets improved wavelet neural network prediction methods urban traffic control long short-term memory neural network traffic engineering computing RFID-ELP data intelligent transportation systems road traffic prediction value term traffic flow prediction approach incorporating vehicle functions short-term traffic flow prediction method RFID systems improved WNN wavelet transforms radio frequency Identification electronic license plate data

来源：评论

学校读者我要写书评

暂无评论

YOLOv4-dense: A smaller and faster YOLOv4 for real-time edge-device based object detection in traffic scene

引用

IET IMAGE PROCESSING 2023年第2期17卷 570-580页

作者： Jiang, Yue Li, Wenjing Zhang, Jun Li, Fang Wu, Zhongcheng Chinese Acad Sci High Magnet Field Lab HFIPS Hefei 230031 Peoples R China Univ Sci & Technol China Hefei Peoples R China High Magnet Field Lab Anhui Prov Hefei Peoples R China

Edge-device-based object detection is crucial in many real-world applications, such as self-driving cars, ADAS, driver behavior analysis. Although deep learning (DL) has become the de-facto approach for object detection, the limited computing resources of embedded devices and the large model size of current DL-based methods increase the difficulty of real-time object detection on edge devices. To overcome these difficulties, in this work a novel YOLOv4-dense model is proposed to detect objects in an accurate, fast manner, which is built on top of the YOLOv4 framework but with substantial improvements. More specifically, lots of CSP layers are pruned since it will decrease inference speed. And to address the losing small objects problem, a dense block is introduced. In addition, a lightweight two-stream YOLO head is also designed to further reduce the computational complexity of the model. Experimental results on NVIDIA JETSON TX2 embedded platform demonstrate that YOLOv4-dense can achieve a higher accuracy, faster speed with smaller model size. For instance, on the KITTI dataset, YOLOv4-dense obtains 84.3% mAP and 22.6 FPS with only 20.3 M parameters, surpassing the state-of-the-art models with comparable parameter budget such as YOLOv3-tiny, YOLOv4-tiny, PP-YOLO-tiny by a large margin.

关键词： traffic engineering computing NVIDIA JETSON TX2 embedded platform feature extraction dense block YOLOv4-dense obtains edge-device-based object detection automobiles driver behavior analysis current DL-based methods smaller YOLOv4 Neural nets YOLOv4-tiny faster YOLOv4 Optical, image and video signal processing Computer vision and image processing techniques convolutional neural nets traffic engineering computing object detection deep learning (artificial intelligence) smaller model size novel YOLOv4-dense model YOLOv4 framework edge devices YOLOv3-tiny embedded devices real-time object detection real-time edge-device computational complexity

来源：评论

学校读者我要写书评

暂无评论

A multi-attention dynamic graph convolution network with cost-sensitive learning approach to road-level and minute-level traffic accident prediction

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第2期17卷 270-284页

作者： Wu, Mingyao Jia, Hongwei Luo, Dan Luo, Haiyong Zhao, Fang Li, Ge Beijing Univ Posts & Telecommun Natl Pilot Software Engn Sch Sch Comp Sci Beijing 100876 Peoples R China Chinese Acad Sci Inst Comp Technol Reseatch Ctr Ubiquitous Comp Syst Beijing Peoples R China Beijing Univ Posts & Telecommun Sch Artificial Intelligence Beijing Peoples R China

traffic accident prediction on road levels and minute levels plays an important role in optimizing public safety and improving traffic infrastructure. However, there are still some challenges in this work. Firstly, the dynamic factors (e.g. traffic flow) affecting traffic accidents make the road network have dynamic spatio-temporal dependency, which leads to biased prediction results. Secondly, the occurrence of traffic accidents is a small probability event, which brings about zero-inflation problem. To address aforementioned problems, the authors propose a Multi-Attention Dynamic Graph Convolution Network with Cost Sensitive Learning approach (MADGCN). Specifically, in the spatial dimension, MADGCN calculates the attention scores of different types of dynamic factors through attention mechanism to simulate the different influence degrees of different factors, and models dynamic inter-road spatial correlation through Graph Convolution Network (GCN). In the temporal aspect, MADGCN adaptively models the dynamic temporal correlations through self-attention blocks. In addition, MADGCN improves the loss function based on cost-sensitive learning strategy to increase the cost of false classification of positive samples, so as to accurately mine sparse positive samples. Experimental results on two real-world traffic accident datasets demonstrate the superiority of MADGCN. Compared with the existing state-of-the-art methods, F1-score of MADGCN on two real-world traffic accident datasets is improved by 11.61% and 9.15%, respectively.

关键词： traffic engineering computing dynamic temporal correlations minute levels MADGCN cost-sensitive learning strategy Combinatorial mathematics road safety learning (artificial intelligence) attention mechanism real-world traffic accident datasets dynamic factors public safety Cost Sensitive Learning approach pattern classification Neural nets biased prediction results convolutional neural nets self-attention blocks traffic engineering computing Data handling techniques attention scores road accidents graph theory intelligent transportation systems minute-level traffic accident prediction probability road traffic road-level road network dynamic spatio-temporal dependency traffic accidents improving traffic infrastructure cost-sensitive learning approach models dynamic inter-road spatial correlation road levels multiattention dynamic graph convolution network

来源：评论

学校读者我要写书评

暂无评论

Prospects and challenges of Metaverse application in data-driven intelligent transportation systems

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第1期17卷 1-21页

作者： Njoku, Judith Nkechinyere Nwakanma, Cosmas Ifeanyi Amaizu, Gabriel Chukwunonso Kim, Dong-Seong Kumoh Natl Inst Technol ICT Convergence Res Ctr Gumi South Korea Kumoh Natl Inst Technol Dept IT Convergence Engn Gumi South Korea

The Metaverse is a concept used to refer to a virtual world that exists in parallel to the physical world. It has grown from a conceptual level to having real applications in virtual reality games. The applicability of the Metaverse in numerous sectors like marketing, education, social, and even advertising exists. However, there exists little or no work on Metaverse applicability to the transportation industry. Data-driven intelligent transportation systems (DDITS) aim to provide more intelligent systems based on exploiting data. This paper reviews the concepts and features of the Metaverse. Also, the review goes over three dominant DDITS challenges: vehicle fault detection and repair, testing new technologies, and anti-theft systems. In addition, it highlights prospective Metaverse solutions that apply to the DDITS. Buttressing the utility of Metaverse in DDITS, this paper presents two major case studies: the invisible to visible (I2V) and the Metaverse on Wheels (MoW) technologies. Finally, the influence, limitations, and open issues of Metaverse applications to DDITS are discussed.

关键词： fault diagnosis transportation Virtual reality virtual world intelligent systems virtual reality data-driven intelligent transportation systems blockchains Computer games exploiting data traffic engineering computing transportation industry dominant DDITS challenges prospective Metaverse solutions Data security intelligent transportation systems virtual reality games computer network security Metaverse applicability Distributed databases physical world Metaverse application Computer communications anti-theft systems

来源：评论

学校读者我要写书评

暂无评论

Ride-pooling in the light of COVID-19: Determining spatiotemporal demand characteristics on the example of MOIA

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第6期17卷 1166-1181页

作者： Zwick, Felix Fraedrich, Eva Axhausen, Kay W. Swiss Fed Inst Technol Inst Transport Planning & Syst Zurich Switzerland MOIA GmbH Hamburg Germany

The mobility provider MOIA operates Europe's largest contiguous electric ride-pooling service in Hamburg, representing a testbed of how shared and digitized transport can help foster the transformation of urban mobility. The on-demand service has been in operation since 2019 and was thus affected by the COVID-19 pandemic in 2020. This study shows real-world insights into travel behavior before and during the pandemic, contributing to the empirical evidence on recent mobility behavior. After the application of descriptive statistical analyses, several (spatial) regression models are estimated to understand the relationship between spatial variables and demand. MOIA trip data from three different time periods are used: (a) before the COVID-19 pandemic in summer and autumn 2019, (b) during the time of the first lockdown in Germany in spring 2020, and (c) after the first lockdown in summer and autumn 2020. A significant positive effect on ride-pooling demand is observed for number of inhabitants, workplaces, gastronomic facilities, and at the airport in all time periods. In the course of the pandemic, the main travel patterns remained stable. However, the positive influences of gastronomy and the airport on ride-pooling demand diminished in 2020. In contrast, the impact of hospitals on ride-pooling demand increased in the course of the pandemic. In areas with high car ownership, ride-pooling demand declined compared to pre-pandemic times.

关键词： Hamburg traffic engineering computing Europe transportation regression models COVID-19 pandemic on-demand service mobility provider traffic engineering computing diseases spatiotemporal demand characteristics digitized transport epidemics regression analysis MOIA trip data Regression analysis urban mobility electric ride-pooling service

来源：评论

学校读者我要写书评

暂无评论

Cognition-inspired behavioural feature identification and motion planning ways for human-like automated driving vehicles

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第4期17卷 754-766页

作者： Xie, Shanshan Zheng, Jingyue Wang, Jianqiang Tsinghua Univ State Key Lab Automot Safety & Energy Beijing 100084 Peoples R China Tsinghua Univ Dept Ind Engn Beijing Peoples R China

Human-like automated driving strategies could have advantages in traffic safety and comfort. However, the primary features of human-like driving behaviors are not clear yet. To deal with this problem, inspired by drivers' cognition way, a simple method is proposed to identify the critical parameters for human-like driving behaviors in extensive scenarios. Then with these parameters as terminal constraints, an interpretable motion planning method is developed in which the longitudinal and lateral planning units are coupled by some state information. With the driving data in the published literature, it is validated that the critical parameters match drivers' behaviors well. With experiments on the simulation and real-car platforms, it is validated that the motion planning method can generate diverse human-like behaviors in multiple scenarios.

关键词： traffic engineering computing traffic safety critical parameters road safety automobiles lateral planning units motion planning ways driving behaviors Mobile robots automated driving vehicles path planning primary features mobile robots longitudinal planning units driver information systems traffic engineering computing drivers driving data Road-traffic system control road traffic automated driving strategies interpretable motion planning method road vehicles Spatial variables control

来源：评论

学校读者我要写书评

暂无评论

Multiple auxiliary classifiers GAN for controllable image generation: Application to license plate recognition

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第1期17卷 243-254页

作者： Shvai, Nadiya Abul Hasnat Nakib, Amir Cydope Ai VINCI Autoroutes Paris France Univ Paris Est Creteil Lab LISSI Vitry Sur Seine France

One of the main challenges in developing machine learning (ML) applications is the lack of labeled and balanced datasets. In the literature, different techniques tackle this problem via augmentation, rendering, and over-sampling. Still, these methods produce datasets that appear less natural, exhibit poor balance, and have less variation. One potential solution is to leverage the Generative Adversarial Network (GAN) which achieves remarkable results in the generation of high-fidelity natural images. However, expanding the ability of GANs' to control generated image attributes with supervisory information remains a challenge. This research aims to propose an efficient method to generate high-fidelity natural images with total control of its main attributes. Therefore, this paper proposes a novel Multiple Auxiliary Classifiers GAN (MAC-GAN) framework based on Auxiliary Classifier GAN (AC-GAN), multi-conditioning, Wasserstein distance, gradient penalty, and dynamic loss. It is therefore presented as an efficient solution for highly controllable image synthesis red that allows to enrich and re-balance datasets beyond data augmentation. Furthermore, the effectiveness of MAC-GAN images on a target ML application called Automatic License Plate Recognition (ALPR) under limited resource constraints is probed. The improvement achieved is over 5% accuracy, which is mainly due to the ability of the MAC-GAN to create a balanced dataset with controllable synthesis and produce multiple (different) images with the same attributes, thus increasing the variation of the dataset in a more elaborate way than data augmentation techniques.

关键词： traffic engineering computing total control auxiliary classifier GAN image recognition data augmentation techniques machine learning applications learning (artificial intelligence) multiple auxiliary classifiers GAN framework MAC-GAN images controllable synthesis pattern classification automatic license plate recognition Neural nets generative adversarial network AC-GAN Computer vision and image processing techniques generated image attributes neural nets traffic engineering computing Data handling techniques Image recognition controllable image generation highly controllable image synthesis red balanced dataset ALPR high-fidelity natural images data augmentation labeled balanced datasets target ML application

来源：评论

学校读者我要写书评

暂无评论

Blinded windows and empty driver seats: The effects of automated vehicle characteristics on cyclists' decision-making

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第1期17卷 72-84页

作者： Bazilinskyy, Pavlo Dodou, Dimitra Eisma, Yke Bauke Vlakveld, Willem de Winter, Joost Delft Univ Technol Dept Cognit Robot Delft Netherlands Delft Univ Technol Dept BioMech Engn Delft Netherlands SWOV The Hague Netherlands

Automated vehicles (AVs) may feature blinded (i.e. blacked-out) windows and external human-machine interfaces (eHMIs), and the driver may be inattentive or absent, but how these features affect cyclists is unknown. In a crowdsourcing study, participants viewed images of approaching vehicles from a cyclist's perspective and decided whether to brake. The images depicted different combinations of traditional vehicles versus AVs, eHMI presence, vehicle approach direction, driver visibility/window-blinding, visual complexity of the surroundings, and distance to the cyclist (urgency). The results showed that the eHMI and urgency level had a strong impact on crossing decisions, whereas visual complexity had no significant influence. Blinded windows caused participants to brake for the traditional vehicle. A second crowdsourcing experiment aimed to clarify the findings of Experiment 1 by also requiring participants to detect the vehicle features. It was found that the eHMI 'GO' and blinded windows yielded high detection rates and that driver eye contact caused participants to continue pedalling. To conclude, blinded windows increase the probability that cyclists brake, and driver eye contact stimulates cyclists to continue cycling. Our findings, which were obtained with large international samples, may help elucidate how AVs (in which the driver may not be visible) affect cyclists' behaviour.

关键词： human-machine interfaces human computer interaction automated vehicle characteristics road safety Mobile robots user interfaces cyclists brake cyclist driver eye contact stimulates cyclists driver seats AV Regression analysis mobile robots Control engineering computing vehicle features traditional vehicle driver information systems traffic engineering computing eHMI presence Road-traffic system control eHMI GO regression analysis blinded windows probability urgency level crowdsourcing decision making human factors User interfaces vehicle approach direction road vehicles large international samples

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：