Communications in Transportation Research最新文献_第3页

Few-shot learning for novel object detection in autonomous driving 基于少镜头学习的自动驾驶新目标检测

IF 12.5

Communications in Transportation Research Pub Date : 2025-07-16 DOI: 10.1016/j.commtr.2025.100194

Yifan Zhuang , Pei Liu , Hao Yang , Kai Zhang , Yinhai Wang , Ziyuan Pu

{"title":"Few-shot learning for novel object detection in autonomous driving","authors":"Yifan Zhuang , Pei Liu , Hao Yang , Kai Zhang , Yinhai Wang , Ziyuan Pu","doi":"10.1016/j.commtr.2025.100194","DOIUrl":"10.1016/j.commtr.2025.100194","url":null,"abstract":"<div><div>Artificial intelligence and advanced sensing technologies have significantly advanced the intelligent transportation system and autonomous vehicles. Perception, a critical component, extracts real-time traffic information essential for various system functionalities, such as agent behavior prediction. However, the quality of information derived from perception greatly influences overall system performance. This study focuses on enhancing perception robustness in autonomous vehicles, particularly in detecting rare objects, which pose a challenge due to limited training samples. While deep learning-based vision methods have shown promising accuracy, they struggle with rare object detection. To address this, we propose a few-shot learning training strategy tailored for improved detection accuracy of rare or novel objects. Additionally, we design a one-stage object detector for efficient object detection in autonomous driving scenarios. Experiments on a self-driving dataset augmented with rare objects alongside the popular few-shot object detection (FSOD) benchmark, the pattern analysis, statical modeling, and computational learning PASCAL Visual Object Classes (PASCAL-VOC), demonstrate state-of-the-art accuracy in rare categories and superior inference speed compared to alternative algorithms. Furthermore, we investigate the impact of intra-class variance on detection accuracy, providing insights for data annotation in the preparation stage.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100194"},"PeriodicalIF":12.5,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144633711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Urban visual clusters and road transport fatalities: A global city-level image analysis 城市视觉集群与道路交通死亡：全球城市级图像分析

IF 12.5

Communications in Transportation Research Pub Date : 2025-07-03 DOI: 10.1016/j.commtr.2025.100193

Zhuangyuan Fan, Becky P.Y. Loo

{"title":"Urban visual clusters and road transport fatalities: A global city-level image analysis","authors":"Zhuangyuan Fan, Becky P.Y. Loo","doi":"10.1016/j.commtr.2025.100193","DOIUrl":"10.1016/j.commtr.2025.100193","url":null,"abstract":"<div><div>Road traffic crashes are among the leading causes of death and injury worldwide. While urban planning and design are known to influence road safety, it is not clear how features of the built environment contribute to traffic fatalities. In this study, we analyze road fatality data from 106 cities across six continents via a combination of computer vision and unsupervised clustering on 26.8 million Google Street View images. We use deep learning tools to extract 25 features from the images. Among these features, 19 are relatively static built environment features, and 6 are dynamic usage-related features (such as pedestrians, cars, buses, and bikes). On the basis of the built environment features, we group the urban streetscapes into six distinct visual clusters. We then examine how these clusters are related to city-level road fatality rates when various control variables (e.g., population size, carbon emissions, income, road length, road safety policy, and continent) and dynamic features are combined. Our findings show that cities with Open Arterials streetscape (extensive road surface, open-sky views, and railings) tend to have higher road fatality rates. After accounting for differences in the built environment, cities with better public transit (proxied by buses detected) tend to have fewer traffic deaths—specifically, a 1% increase in bus presence is linked to a 0.35% decrease in fatalities per 100,000 people. This study demonstrates the power of using widely available street view imagery to uncover global disparities in urban design and their connection to road safety.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100193"},"PeriodicalIF":12.5,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144535777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SAFER-predictor: Sparse adversarial training framework for robust traffic prediction under missing and noisy data SAFER-predictor：稀疏对抗训练框架，用于缺失和噪声数据下的鲁棒交通预测

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-26 DOI: 10.1016/j.commtr.2025.100192

Yutian Liu , Chengfeng Jia , Soora Rasouli , Jian Gong , Tao Feng , Melvin Wong , Tianjin Huang

{"title":"SAFER-predictor: Sparse adversarial training framework for robust traffic prediction under missing and noisy data","authors":"Yutian Liu , Chengfeng Jia , Soora Rasouli , Jian Gong , Tao Feng , Melvin Wong , Tianjin Huang","doi":"10.1016/j.commtr.2025.100192","DOIUrl":"10.1016/j.commtr.2025.100192","url":null,"abstract":"<div><div>Accurate traffic flow forecasting is essential for developing intelligent transportation systems (ITSs) to reduce congestion, optimize road management, and improve safety. While data-driven traffic prediction approaches have shown high accuracy, they rely heavily on precise measurements, making them vulnerable to perturbed environmental factors, like sensor malfunctions, data storage issues, and adverse weather conditions. To overcome the limitation, we propose SAFER-Predictor, a novel sparse adversarial training (Sparse AT) framework for enhancing the reliability of deep learning based spatiotemporal traffic prediction models. Sparse AT extends traditional adversarial training (AT) through a two-phase process: pre-training and fine-tuning. In the pre-training phase, the model is optimized to capture normal traffic patterns, enhancing predictive performance by understanding standard dynamics without external disruptions. In the fine-tuning phase, the focus shifts to strengthening robustness against corrupted inputs by employing an iterative min-max strategy during AT, optimizing performance for worst-case scenarios. Furthermore, we derive theoretical formulations that establish an upper bound on the model's prediction error following Sparse AT under certain noise levels. Experimental results indicate that incorporating Sparse AT into the representative traffic flow prediction models improves stability and ensures high accuracy under various perturbation scenarios.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100192"},"PeriodicalIF":12.5,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Customized recursive model for drivers’ navigation compliance behaviors under abnormal events 自定义异常事件下驾驶员导航遵从行为递归模型

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-23 DOI: 10.1016/j.commtr.2025.100187

Kaijie Zou, Yaming Guo, Ke Zhang, Meng Li

{"title":"Customized recursive model for drivers’ navigation compliance behaviors under abnormal events","authors":"Kaijie Zou, Yaming Guo, Ke Zhang, Meng Li","doi":"10.1016/j.commtr.2025.100187","DOIUrl":"10.1016/j.commtr.2025.100187","url":null,"abstract":"<div><div>In recent years, the resilience of road traffic during abnormal events has drawn considerable attention. Intelligent navigation systems, which proactively guide drivers along optimal routes in such situations, are viewed as a promising solution to facilitate recovery of road network performance. A key question arises: How do drivers choose routes when guided by navigation systems? This study addresses that question by modeling drivers’ decision-making behavior at each decision point using a nested framework. At the upper level, drivers decide whether to strictly follow the route recommended by the navigation system, while at the lower levels, they make route choices in the absence of guidance. A Customized Nested Dynamic Recursive Logit (C-NDRL) model was developed to capture these behaviors. Parameters for both decision levels were jointly estimated using a Broyden-Fletcher-Goldfarb-Shanno (BFGS) Method-based algorithm, and the model was verified on the Sioux-Falls network. The model was then applied to real navigation route and driving trajectory data from Canton, China, for parameter estimation and the analysis of the additional utility provided by navigation. The results indicate that the C-NDRL model significantly outperformed other models. Furthermore, the study quantifies the substantial impact of external environmental factors and navigation-related internal factors on drivers’ compliance on navigation systems, highlighting that during rainstorm days, the additional utility from navigation increases by 17%.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100187"},"PeriodicalIF":12.5,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144364486","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards robust motion control in multi-source uncertain scenarios by robust policy iteration 基于鲁棒策略迭代的多源不确定场景鲁棒运动控制研究

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-20 DOI: 10.1016/j.commtr.2025.100191

Jie Li , Letian Tao , Wenjun Zou , Yuhang Zhang , Bin Shuai , Jingliang Duan , Shengbo Eben Li , Hao Sun , Yiru Wang , Yu Gao , Yuwen Heng , Anqing Jiang

{"title":"Towards robust motion control in multi-source uncertain scenarios by robust policy iteration","authors":"Jie Li , Letian Tao , Wenjun Zou , Yuhang Zhang , Bin Shuai , Jingliang Duan , Shengbo Eben Li , Hao Sun , Yiru Wang , Yu Gao , Yuwen Heng , Anqing Jiang","doi":"10.1016/j.commtr.2025.100191","DOIUrl":"10.1016/j.commtr.2025.100191","url":null,"abstract":"<div><div>The adoption of neural networks for motion control modules emerges as a critical direction in the advancement of end-to-end autonomous driving. However, few studies have comprehensively addressed the challenges of robustness and generalization in motion control policies, including long-tailed distribution, distribution shift, and sim-to-real gap. In practical applications, motion control performance is compromised by diverse uncertainties, posing substantial challenges to real-world deployment. This work develops a training system to enhance the robustness and generalization of motion control policies when passing through multiple intersections. We first construct a task library comprising 6 driving scenarios, which are allocated to different sampling processes to rebalance the proportion of monotonous and edge scenarios. Next, we formulate a zero-sum game for uncertainties and driving actions with smoothing constraints within the range of observation noise. The driving policy is optimized by the proposed robust policy iteration method for the worst-case performance, which is approximated via Taylor expansion to avoid the computational burden caused by adversarial training on behavior disturbance, where the approximate results decouple model mismatches to ensure robust performance and action smoothness is boosted through penalty function method. Ultimately, the motion control performance and the robustness of driving policy are thoroughly validated by configuring the behavior patterns of traffic participants, ego dynamic parameters, and observation noise intensities in the simulation environment. Physical vehicle experiments on public urban roads further depict the robustness and generalization of the driving policy learned from simulations.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100191"},"PeriodicalIF":12.5,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144330317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A sustainable multi-objective framework for multi-phased, capacitated vertiport siting with land use integration 一个可持续的多目标框架，用于多阶段、有能力的垂直机场选址和土地利用整合

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-18 DOI: 10.1016/j.commtr.2025.100186

Hao Wu , Shahriar Iqbal Zame , Tao Guo , Qing-Long Lu , Constantinos Antoniou

{"title":"A sustainable multi-objective framework for multi-phased, capacitated vertiport siting with land use integration","authors":"Hao Wu , Shahriar Iqbal Zame , Tao Guo , Qing-Long Lu , Constantinos Antoniou","doi":"10.1016/j.commtr.2025.100186","DOIUrl":"10.1016/j.commtr.2025.100186","url":null,"abstract":"<div><div>This research presents a multi-objective optimization framework for incremental siting of capacitated vertiports that integrates real land use data and aims to maximize generalized cost savings while minimizing infrastructure costs and emissions. The multi-phased siting framework uniquely facilitates the gradual evolution of Urban Air Mobility (UAM) operations from initial electric vertical takeoff and landing vehicles (eVTOLs) to more advanced modular flying vehicles (MFVs). This phased technological progression provides a practical pathway toward fully operational flying cars while ensuring feasible infrastructure adaptability across these transitions. Applied to the Munich metropolitan area, the framework demonstrates that multi-phased siting, particularly a 4-phased strategy, yielding about 1.315 × 10<sup>5</sup> euros higher daily net profits. Specifically, compared to base single-phased approach, the 4-phased strategy delivers substantial marginal improvements across key metrics for an exemplary operating day: 1.3 × 10<sup>4</sup> euros in generalized travel cost savings, 15 t in emissions reductions, and a 0.9% increase in UAM mode share. Beyond four phases, the benefits diminish relative to increased complexity. A full factorial analysis examining capacity constraints and infrastructure costs reveals that ignoring either factor leads to impractical outcomes-unconstrained capacity results in demand exceeding 60-fold capacity, while disregarding infrastructure costs generates negative net profits due to overinvestment. The analysis identifies an optimal infrastructure cost subsidy range of 20%–40%, balancing performance gains with economic sustainability. These findings enable integrated planning that effectively balances operational efficiency, system-wide environmental externalities, and economic viability through optimized cost allocation and phased investment strategies.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100186"},"PeriodicalIF":12.5,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144312865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Situational awareness using set-based estimation and vehicular communication: An occluded pedestrian-crossing scenario 使用基于集合的估计和车辆通信的态势感知：一个闭塞的行人过街场景

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-11 DOI: 10.1016/j.commtr.2025.100190

Vandana Narri , Amr Alanwar , Jonas Mårtensson , Henrik Pettersson , Fredrik Nordin , Karl Henrik Johansson

引用次数: 0

DFUN-KDF: Efficient and robust decentralized federated framework for UAV networks via knowledge distillation and filtering DFUN-KDF：基于知识蒸馏和过滤的高效鲁棒的去中心化无人机网络联邦框架

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-11 DOI: 10.1016/j.commtr.2025.100173

Wenyuan Yang , Yuhang Liu , Xinlin Leng , Hanlin Gu , Gege Jiang , Xiaochuan Yu , Xiaochun Cao

{"title":"DFUN-KDF: Efficient and robust decentralized federated framework for UAV networks via knowledge distillation and filtering","authors":"Wenyuan Yang , Yuhang Liu , Xinlin Leng , Hanlin Gu , Gege Jiang , Xiaochuan Yu , Xiaochun Cao","doi":"10.1016/j.commtr.2025.100173","DOIUrl":"10.1016/j.commtr.2025.100173","url":null,"abstract":"<div><div>Unmanned aerial vehicles (UAVs) are increasingly crucial across various fields. There is a growing interest in using federated learning (FL) methods to enhance the efficiency of UAV operations. Nevertheless, incumbent methods remain encumbered by significant drawbacks, including high energy consumption from extensive parameter exchanges, the imperative for homogeneous networks, and sensitivity to single-point failures. These difficulties are compounded by the unreliable nature of communication channels and the current inability to effectively manage the diversity of UAV models, highlighting the imperative for more resilient and adaptable FL solutions. To address these issues, we propose an efficient and robust decentralized FL framework for heterogeneous UAV networks. Our framework first leverages the knowledge distillation where UAVs transmit embeddings instead of model parameters to reduce the number of transmission parameter. UAVs update their local models using embeddings generated by other UAVs, which also enables UAVs with diverse architectures to participate in training. Moreover, our framework incorporates a filtering mechanism to remove malicious embeddings, ensuring resilience against adversities in UAV networks. Extensive experiments on various datasets validate the effectiveness and practical deployment potential of our framework.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100173"},"PeriodicalIF":12.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144262802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integrating micro and macro traffic control for mixed autonomy traffic 混合自治交通宏微观一体化控制

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-10 DOI: 10.1016/j.commtr.2025.100188

Tingting Fan , Jieming Chen , Edward Chung

{"title":"Integrating micro and macro traffic control for mixed autonomy traffic","authors":"Tingting Fan , Jieming Chen , Edward Chung","doi":"10.1016/j.commtr.2025.100188","DOIUrl":"10.1016/j.commtr.2025.100188","url":null,"abstract":"<div><div>During the transition to fully autonomous traffic systems, managing mixed traffic consisting of connected automated vehicles (CAVs) and human-driven vehicles (HDVs) is imperative. Existing macroscopic and microscopic strategies have shown effectiveness in alleviating highway congestion. However, the integration of these strategies for mixed autonomy traffic remains under-explored. This study proposes a hybrid flow and trajectory control (HFTC) strategy that combines a macroscopic control, ramp metering (RM), with a microscopic control, cooperative merging (CM) for CAV trajectory optimization in mixed traffic scenarios. Specifically, the RM control considers CAV-penetration-dependent dynamics to regulate ramp flow, and the CM utilizes a centralized optimization model to enhance CAV merging trajectories. Independently implementing RM or CM proved effective only under heavy or moderate traffic flow, whereas our proposed integrated strategy, HFTC, demonstrated greater adaptability and suitability under various traffic conditions. Additionally, the impacts of CAV penetration rates and traffic flows on performance of different control strategies are thoroughly explored. Simulation results indicate that under low and moderate traffic conditions, microscopic control can be comparable to macroscopic control given sufficient CAV integration, while under heavy traffic flows, macroscopic control cannot be replaced by microscopic control.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100188"},"PeriodicalIF":12.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A multiagent social interaction model for autonomous vehicle testing 自动驾驶汽车测试中的多智能体社会交互模型

IF 12.5

Communications in Transportation Research Pub Date : 2025-06-10 DOI: 10.1016/j.commtr.2025.100183

Shihan Wang , Ying Ni , Chengsheng Miao , Jian Sun , Jie Sun

{"title":"A multiagent social interaction model for autonomous vehicle testing","authors":"Shihan Wang , Ying Ni , Chengsheng Miao , Jian Sun , Jie Sun","doi":"10.1016/j.commtr.2025.100183","DOIUrl":"10.1016/j.commtr.2025.100183","url":null,"abstract":"<div><div>Social interaction capability (SIC) is essential for autonomous vehicles (AVs) when they interact with surrounding vehicles, as the ability of understanding and reacting to the behaviors of other road users can significantly enhance AVs’ rapid deployment. Virtual simulation testing is a core approach for evaluating AVs, including their SIC, on the basis of traffic simulation models. However, existing simulation models focus mainly on generating accurate vehicle trajectories and do not explicitly model the high-level sociality nature of interaction decisions that guide specific movements. This study aims to address this gap by developing a multiagent simulation model for the social interaction of human driving behavior on the basis of the multiagent imitation learning (MAIL) approach, which is referred to as the Social-MAIL model. Specifically, to quantify the sociality of decisions, we introduce social value orientation into the reward function to quantify cooperation or competition intent and guide the generation of social driving behaviors. Furthermore, to fully depict the complex interaction environment, we develop a heterogeneous policy network with temporal‒spatial attention mechanisms to describe the impact of multiple interactive objects and historical states on driving behavior. Through training and validation on the SinD dataset, we demonstrate that, compared with a set of baseline models, the proposed Social-MAIL model can accurately capture complex and time-varying social intent and reproduce the most realistic vehicle trajectories and macroscopic traffic flow characteristics at intersections. Moreover, we apply the Social-MAIL model for evaluating the SIC of AVs via comparison experiments.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100183"},"PeriodicalIF":12.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0