IEEE Transactions on Intelligent Transportation Systems最新文献

{"title":"Self-Supervised Transformer for Trajectory Prediction Using Noise Imputed Past Trajectory","authors":"Vibha Bharilya;Ashok Arora;Neetesh Kumar","doi":"10.1109/TITS.2025.3550711","DOIUrl":"https://doi.org/10.1109/TITS.2025.3550711","url":null,"abstract":"Trajectory prediction is one of the important components for achieving higher levels of Society of Automotive Engineers (SAE) driving automation, enabling them to navigate through complex driving scenarios and make informed decisions in unexplored roads. Problems such as the varying behaviour of drivers on the road, sensor measurement inaccuracies, and dense and complex environmental circumstances make this task difficult. To address these challenges, a self-supervised transformer (SST) framework is proposed. The noise-imputed trajectory points for road agents are generated. This enhances the model’s ability to handle uncertain data. A self-supervised task is proposed, which focuses on predicting past trajectory points using both true and noise-imputed trajectory encoded features. This approach highlights the important patterns or connections in the data that could go unnoticed if supervised tasks were the only one used. Further, in addition to the trajectory prediction task, a consistent loss function should be introduced to preserve spatial consistency with noise imputed trajectory points. Moreover, learnable query embeddings are added to the system to improve the diversity of multi-modal predictions. The SST model has been evaluated on the widely used and recent Argoverse 2 dataset and outperforms state-of-the-art models by a margin of 2.17%-24.15%, 5.38%-30.05%, 9.41%-46.89% and 0.20%-21.21% on the minADE6, minFDE6, MR6 and b-FDE6 respectively.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"8454-8466"},"PeriodicalIF":7.9,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196596","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Secure Consensus for Multi-Agent Systems With Euler-Lagrange Dynamics and Multiple DoS Attacks","authors":"Ruimei Zhang;Liang Liu;Ju H. Park;Deqiang Zeng;Xiangpeng Xie","doi":"10.1109/TITS.2025.3554274","DOIUrl":"https://doi.org/10.1109/TITS.2025.3554274","url":null,"abstract":"This article is focused on the secure consensus of multi-agent systems (MASs) with Euler-Lagrange (EL) dynamics and multiple DoS attacks. First, a new model of the multiple DoS attacks is built, which is based on discrete sampled-data communication and considers the joint impact of the multiple DoS attacks. Second, under multiple DoS attacks, two new technical results are proposed, which lay a good foundation for consensus analysis. Then, by designing an adaptive distributed control (DC) protocol combining with two new auxiliary systems, secure consensus results are derived for MASs with EL dynamics and multiple DoS attacks. Finally, simulations based on networked two-linked robot manipulators are presented to show the effectiveness of the theoretical results.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 7","pages":"10008-10018"},"PeriodicalIF":7.9,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"UnMoDE: Uncertainty Modeling for Driver Gaze Estimation via Feature Disentanglement","authors":"Daosong Hu;Mingyue Cui;Kai Huang","doi":"10.1109/TITS.2025.3556553","DOIUrl":"https://doi.org/10.1109/TITS.2025.3556553","url":null,"abstract":"Gaze estimation can be used for assessing the attention level of drivers. Current works predominantly focus on enhancing model accuracy, often overlooking the influence of input sample and label uncertainty. In this paper, we propose a framework for uncertainty modeling in driver gaze estimation via feature disentanglement, referred to as UnMoDE. Our approach begins by extracting facial information into distinct feature spaces using an asymmetric dual-branch encoder to obtain gaze features. Subsequently, a multi-layer perceptron (MLP) is employed to project gaze features and labels into an embedding space, representing them as Gaussian distributions. The uncertainty is described using a covariance matrix. Random sampling is applied to derive samples from the gaze embedding distribution to estimate the most probable embedding representation. This estimated representation is then used to regress the gaze direction and is projected back into the gaze feature space, along with identity information, to facilitate facial reconstruction. Extensive experimental evaluations demonstrate that UnMoDE significantly outperforms baseline and state-of-the-art methods on the latest benchmark datasets collected for drivers, particularly in reducing the number of samples with significant errors.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 7","pages":"10612-10622"},"PeriodicalIF":7.9,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Robust Tensor Ring Decomposition for Urban Traffic Data Imputation","authors":"Linfang Yu;Chenyu Guan;Hao Wang;Yuxin He;Wenming Cao;Chi-Sing Leung","doi":"10.1109/TITS.2025.3555449","DOIUrl":"https://doi.org/10.1109/TITS.2025.3555449","url":null,"abstract":"In urban transportation systems, missing data and noise contamination are almost inevitable. To address the challenge of imputing traffic data corrupted by noise and outliers in real-world scenarios, this paper proposes a novel algorithm based on spatiotemporal tensor completion. The proposed method transforms observed data into three-dimensional spatiotemporal tensors and utilizes tensor ring decomposition for data completion. Furthermore, spatial and temporal information is incorporated into the model by utilizing the graph Laplacian matrix. To handle outliers, they are treated as unknown parameters, and the <inline-formula> <tex-math>$ell _{0}$ </tex-math></inline-formula>-norm is introduced to ensure their sparsity, thereby achieving the Spatio-Temporal Tensor Completion model with <inline-formula> <tex-math>$ell _{0}$ </tex-math></inline-formula>-norm term (STTC-<inline-formula> <tex-math>$ell _{0}$ </tex-math></inline-formula>). The solution to the model is derived using the alternating optimization framework with the alternating direction method of multipliers. Then, we discuss the convergence of the solution method. To further enhance the efficiency of our proposed method, we combine the unrolling algorithm with our iterative optimization model, creating a lightweight and efficient neural network tailored for tensor completion, called STTC-<inline-formula> <tex-math>$ell _{0}$ </tex-math></inline-formula>-NN. Extensive experiments conducted on real datasets demonstrate the superiority of our proposed method over several state-of-the-art methods across various experimental scenarios. It is worth noting that STTC-<inline-formula> <tex-math>$ell _{0}$ </tex-math></inline-formula>-NN reduces computational time by one to two orders of magnitude compared to existing methods while maintaining or even improving imputation accuracy. The code is available at <uri>https://github.com/TCCofWANG/STTC-L0-and-STTC-L0-NN</uri>.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"8707-8719"},"PeriodicalIF":7.9,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Attentive Radiate Graph for Pedestrian Trajectory Prediction in Disconnected Manifolds","authors":"Peiyuan Zhu;Shengjie Zhao;Hao Deng;Fengxia Han","doi":"10.1109/TITS.2025.3555390","DOIUrl":"https://doi.org/10.1109/TITS.2025.3555390","url":null,"abstract":"Pedestrian trajectory prediction grapples with the demanding feat of modeling complex interactions and learning multimodal distribution to navigate different human-centric environments. Despite superior performance in reducing distance-based metrics, recent works tend to predict out-of-distribution trajectories, as the distribution of forthcoming paths comprises a blend of various manifolds that may be disconnected. These unrealistic trajectories can potentially jeopardize the safety of traffic participants and result in significant damage. To meet these challenges, we propose DMPred, a graph-based generator adversarial network that generates realistic multimodal trajectory predictions by better modeling the social interactions of pedestrians across different scenes in disconnected manifolds. The core of DMPred is an attentive radiate graph sequence constructed by considering the localized influence radiating from pedestrian movements, which is followed by a spatiotemporal extractor that stores and reuses potentially forgotten neighboring pedestrian information to allow for better extraction of complex interactions. Additionally, a collection of generators is utilized for forecasting, which incorporates spectral clustering on trajectories during the prior learning process of multiple generators to help reduce model redundancy and enhance flexibility for various prediction scenarios. Through extensive experiments on multiple real-world and simulation datasets, we demonstrate that DMPred obtains highly competitive results with efficacy in predicting realistic multimodal trajectories.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"7755-7769"},"PeriodicalIF":7.9,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144206141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual Cyber-Physical Network-Based Optimization of Cloudlet Formation for the Internet of Vehicles","authors":"Yun Meng;Fan Liu;Xinyi Liu;Wei Wang;Liang Dai","doi":"10.1109/TITS.2025.3554847","DOIUrl":"https://doi.org/10.1109/TITS.2025.3554847","url":null,"abstract":"The vehicular cloudlet (VC), capable of leveraging synergies to support emerging cooperative services and task offloading among neighboring vehicles, will be adopted in the Internet of Vehicles (IoV). Compared with traditional clustering methods based on link connectivity, VCs require higher transmission capacity and stability. Three major challenges must be addressed when optimally establishing the VC structure. First, the transmission capacity is affected by inherent stochastic characteristics, including channel fading and interference. Second, the mobility of vehicles introduces instability. Third, a comprehensive optimization model is required to jointly improve the stability and transmission capacity of the established VC. Therefore, this paper proposes a dual cyber-physical network (DCP) model to represent the dynamic physical network and the coupled transmission network. Generalized analytical expressions for channel quality are derived using moment-generating functions based on a Nakagami-m small-scale fading model to handle stochastic characteristics. Furthermore, a DCP association density optimization model is proposed that considers the stability of the physical topology and the transmission capacity of the channel. Symmetric non-negative matrix factorization is used to solve the optimization problem with low complexity. Simulation results confirm that our proposed method achieves higher transmission capacity and stability compared existing link connectivity-based clustering methods.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"7486-7495"},"PeriodicalIF":7.9,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Waterway-BEV: Generate Bird’s Eye View Layouts of a Waterway From a First-Person View Camera Using Cross-View Transformers","authors":"Feng Ma;Xin Jiang;Chen Chen;Jie Sun;Xin-Ping Yan;Jin Wang","doi":"10.1109/TITS.2025.3554717","DOIUrl":"https://doi.org/10.1109/TITS.2025.3554717","url":null,"abstract":"In the domain of autonomous ship navigation, the construction of bird’s-eye view (BEV) layouts for waterways has obvious significance. A helmsman can generate the BEV layout of the waterway using his/her eyes only. To simulate this intelligence, a novel neural network-based algorithm named Waterway-BEV is proposed, which enables reconstructing a local map formed by the waterway layout and ship occupancies in the bird’s-eye view given a first person view monocular image only. Waterway-BEV employs an efficient SEResNeXt encoder to extract features from first person view (FPV) monocular images, capturing deep semantic information related to waterways and ships. Due to the variations in information across different perspectives, Waterway-BEV incorporates a Cross-View Transformation Module, which takes the constraint of cycle consistency between views into account and makes full use of their correlation to strengthen the view transformation and scene understanding. To fully leverage the feature output of the SEResNeXt encoder, Waterway-BEV employs a decoder based on a dedicated lightweight network. This decoder is responsible for decoding the enhanced bird’s-eye view (BEV) feature maps and generating the BEV layout. By employing the Focal Loss as the loss function for model optimization, Waterway-BEV takes into account the quantity and classification difficulty of ship samples during the training process, thereby improving the generation performance and convergence speed. The experiments demonstrated that Waterway-BEV achieved notable performance metrics, with mIOU and mAP rates reaching 97.8% and 98.2%, respectively, in waterway bird’s-eye view layout generation. Waterway-BEV outperformed other state-of-the-art (SOTA) algorithms in generating BEV layouts of waterways. In particular, during specialized scenarios such as crossroads of waterways and tasks involving small target ships, Waterway-BEV consistently generated satisfactory bird’s-eye view layouts, demonstrating robustness and applicability.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"8078-8096"},"PeriodicalIF":7.9,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploratory Driving Performance and Car-Following Modeling for Autonomous Shuttles Based on Field Data","authors":"Renan Favero;Lily Elefteriadou","doi":"10.1109/TITS.2025.3552506","DOIUrl":"https://doi.org/10.1109/TITS.2025.3552506","url":null,"abstract":"Autonomous shuttles (AS) operate in several cities and have shown potential to improve the public transport network. However, there is no car-following model that is based on field data and allows decision-makers (planners, and traffic engineers) to assess and plan for AS operations. To fill this gap, this study collected field data from AS operations, analyzed their driving performance, and suggested changes in the AS trajectory model to improve passenger comfort. A sample was collected with over 4,000 seconds of data of AS following a conventional car (human driver). The sample contained GPS positions from both AS and conventional vehicles. Latitude and longitude coordinates were used to calculate the speed, acceleration, and jerk of the leader and follower. The data analyses indicated that AS has higher jerk values that may impact the passengers’ comfort. Several existing models were evaluated, and the researchers concluded that the calibrated ACC model resulted in lower errors for AS spacing and speed. The results of the calibration indicate that the AS exhibits lower peak acceleration and higher deceleration than those found in calibrated parameters of autonomous vehicle models from other studies.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 5","pages":"6042-6055"},"PeriodicalIF":7.9,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143908424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CrackVisionX: A Fine-Tuned Framework for Efficient Binary Concrete Crack Detection","authors":"Abdulrahman A. ALKannad;Ahmad Al Smadi;Moeen Al-Makhlafi;Shuyuan Yang;Zhixi Feng","doi":"10.1109/TITS.2025.3546770","DOIUrl":"https://doi.org/10.1109/TITS.2025.3546770","url":null,"abstract":"Cracks are critical defects in concrete structures, traditionally identified through human inspection. However, computer vision techniques, especially convolutional neural networks (CNNs), offer promising solutions for automated detection. Driven by this trend, this study proposes CrackVisionX, a state-of-the-art deep learning framework for classifying binary concrete cracks. CrackVisionX lies in its integration of advanced CNN architectures, ResNet50, MobileNet_v3_large, DenseNet121, and EfficientNetB0, with extensive hyper-parameter tuning. This integration optimizes crack detection accuracy while maintaining low model complexity and reducing bias, making it suitable for real-time applications. Furthermore, the framework introduces a robust data augmentation strategy that effectively addresses dataset imbalances, enhancing model generalization across diverse domains. Additionally, CrackVisionX employs comprehensive preprocessing on the METU and SDNET2018 datasets to create six domains: Bridge Deck, Wall, Pavement, SDNET2018, METU, and METU & SDNET2018. The framework’s performance is thoroughly evaluated and benchmarked against state-of-the-art methods, utilizing diverse metrics to improve the detection of cracks in concrete structures. EfficientNetB0, a core component of the framework, demonstrated superior performance with exceptional test accuracies of up to 99.71%, 99.78%, 99.55%, 99.89%, 99.98%, and 99.92% for Bridge Deck, Wall, Pavement, SDNET2018, METU, and METU & SDNET2018, respectively. Moreover, we evaluated the robustness of CrackVisionX using images contaminated with different types and intensities of noise, demonstrating its reliability and effectiveness. This balance between high accuracy and computational efficiency confirms the framework’s potential for practical deployment. The experimental results emphasize the transformative potential of deep learning in construction safety and structural health monitoring.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 7","pages":"10353-10372"},"PeriodicalIF":7.9,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536512","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Precision Traffic Monitoring: Leveraging Distributed Acoustic Sensing and Deep Neural Networks","authors":"Yacine Khacef;Martijn van den Ende;Cédric Richard;André Ferrari;Anthony Sladen","doi":"10.1109/TITS.2025.3556234","DOIUrl":"https://doi.org/10.1109/TITS.2025.3556234","url":null,"abstract":"Distributed Acoustic Sensing (DAS) has recently emerged as a promising technology for traffic monitoring. It transforms standard fiber-optic telecommunication cables into an array of vibration sensors capable of capturing vehicle-induced subsurface deformation with high spatio-temporal resolution. In this study, we propose a deep learning framework for the detection and velocity estimation of traffic flow. Our neural network based model yields accurate and well-resolved vehicle localization and speed tracking, outperforming off-the-shelf Dynamic Time Warping based solutions while achieving an order of magnitude faster processing time. A multi-day comparison with dedicated sensors installed along an urban highway shows a strong correlation, even under dense traffic conditions.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 6","pages":"7678-7689"},"PeriodicalIF":7.9,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144205814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}