<rss version="2.0" xmlns:atom="https://www.w3.org/2005/Atom">
  <channel>
    <title>Transport Research International Documentation (TRID)</title>
    <link>https://trid.trb.org/</link>
    <atom:link href="https://trid.trb.org/Record/RSS?s=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" rel="self" type="application/rss+xml" />
    <description></description>
    <language>en-us</language>
    <copyright>Copyright © 2026. National Academy of Sciences. All rights reserved.</copyright>
    <docs>http://blogs.law.harvard.edu/tech/rss</docs>
    <managingEditor>tris-trb@nas.edu (Bill McLeod)</managingEditor>
    <webMaster>tris-trb@nas.edu (Bill McLeod)</webMaster>
    <image>
      <title>Transport Research International Documentation (TRID)</title>
      <url>https://trid.trb.org/Images/PageHeader-wTitle.jpg</url>
      <link>https://trid.trb.org/</link>
    </image>
    <item>
      <title>Real LiDAR point cloud synthesis for 3D object detection in snowy weather</title>
      <link>https://trid.trb.org/View/2680758</link>
      <description><![CDATA[Light Detection And Ranging (LiDAR) sensors can generate a number of sequential 3D point clouds, which are widely deployed in many real-world systems. 3D object detection in point clouds, is one of the most fundamental tasks. Unfortunately, the existing 3D object detection methods degrade in snowy weather, because in that situation the annotated samples are difficult to collect. To solve this issue, we propose a novel GAN-based Snowfall Point-cloud AugmentOR (GANspaor) to generate high-quality synthetic snowfall point clouds as augmentations. The basic idea of GANspaor is to transfer annotated point clouds to snowfall versions by simultaneously learning the global style of real snowfall point clouds and the local details of physics-induced ones. Our framework fuses data-driven and physical modeling methods for rapidly generating data in snowy weather. To evaluate the effectiveness of GANspaor, we employ a number of recent 3D object detection methods and train them by using the synthetic samples of GANspaor as auxiliary augmentations. Moreover, we conduct a comparative analysis of the characteristics of the data distributions of the snowy point clouds synthesized by GANspaor. Experimental results demonstrate that GANspaor can improve the performance of 3D object detection methods compared with other existing snowfall point cloud simulators.]]></description>
      <pubDate>Tue, 30 Jun 2026 10:21:25 GMT</pubDate>
      <guid>https://trid.trb.org/View/2680758</guid>
    </item>
    <item>
      <title>Vision-based modal identification and weight estimation of vehicles</title>
      <link>https://trid.trb.org/View/2684460</link>
      <description><![CDATA[Vehicle weight estimation plays a crucial role in evaluating the fatigue life of bridges and pavements. Traditional Weigh-in-Motion (WIM) systems, which require sensors on bridges or pavements, are costly and have limited deployment flexibility. Recently, computer vision-based approaches have been explored to estimate vehicle weight from vibration characteristics of vehicles without the need for such sensors. However, most existing methods employ simplified vehicle models that ignore damping effects, resulting in limited estimation accuracy due to the inherently high damping ratio of vehicles. Moreover, these methods often rely on prior knowledge of the vehicle’s center of gravity (CoG), further restricting their general applicability.This study proposes an enhanced vision-based framework that overcomes these limitations. Vehicle vibrations are captured through object detection, filtering, and subpixel displacement estimation. The System Realization through Information Matrix (SRIM) method is then used to identify damped bouncing and pitching modes with complex mode shapes. Based on the state-space equations, a complex matrix formulation incorporating these damped modes enables simultaneous estimation of vehicle and axle weights, achieving over 84% accuracy in field tests.]]></description>
      <pubDate>Tue, 30 Jun 2026 10:21:01 GMT</pubDate>
      <guid>https://trid.trb.org/View/2684460</guid>
    </item>
    <item>
      <title>High Adversarial Robustness Network: Adaptive Positional Encoding and Parallel Attention for Obstacle Recognition in Autonomous Driving</title>
      <link>https://trid.trb.org/View/2617923</link>
      <description><![CDATA[Deep neural networks (DNNs) are critical for obstacle recognition in autonomous driving, commonly used to classify objects like vehicles and animals. However, DNNs are vulnerable to adversarial attacks that can cause misclassifications and compromise system safety. To address this, we propose the Adaptive Multi-Scale Positional Encoding Parallel Attention Network (APANet), a model designed to enhance adversarial robustness. APANet includes four main components: multi-scale feature map generation, Adaptive Multi-Scale Positional Encoding (AMSPE), Parallel Attention (PA), and multi-scale feature fusion. AMSPE embeds adaptive positional information and captures long-range dependencies to boost resistance to adversarial perturbations. PA independently processes multi-scale features, enhancing feature utilization and isolating adversarial noise. These components work synergistically to improve the model’s robustness. Experiments show APANet significantly outperforms several state-of-the-art models in Top-1 accuracy under various adversarial attacks and on clean samples. Specifically, AMSPE contributes a 4.13-point improvement in adversarial accuracy and narrows the clean-adversarial performance gap by 4.73 points, while PA improves recognition accuracy by 6.11 points. To validate real-world robustness, we tested APANet on the German Traffic Sign Recognition Benchmark (GTSRB), where adversarial interference can critically affect autonomous driving. APANet demonstrates high accuracy and robustness under adversarial scenarios on GTSRB, confirming its effectiveness in enhancing the safety and reliability of autonomous driving systems.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:14 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617923</guid>
    </item>
    <item>
      <title>Multiple Autonomous Underwater Vehicles-Assisted Data Collection in 6G-Driven Underwater Wireless Networks Based on Software-Defined MARL</title>
      <link>https://trid.trb.org/View/2617907</link>
      <description><![CDATA[The multiple Autonomous Underwater Vehicle (AUV)-assisted cooperative system or the AUV-based Underwater Ad-hoc Networks (UAN) system has been considered as a highly-potential future in underwater data surveillance. In this paper, we propose grid-based distributed data collection architecture and define two categories of navigation modes. Based on the proposed data collection model, we propose MADAC, a scheme based on AUV-based UAN to cooperatively collect data from 6G-driven underwater wireless networks. We utilize the Software-Defined Networking (SDN) technique to re-organize the architecture of AUV-based UAN and propose software-defined actor-critic MARL framework. Based on the proposed MARL framework, we present the paradigm of MADDPG algorithm with optimal similarity attention mechanism (MADDPG-SA), to plan the paths for the AUV-based UAN, especially the cooperative underwater obstacle avoidance, the task distribution balancing, the Value of Information (VoI) are concurrently taken into account. In particular, the proposed MADDPG-SA improves the running efficiency of the proposed MADDPG-SA by encouraging the agent to learn from the similar and better-performance agent. The evaluation results demonstrate that the proposed MADAC can schedule the AUV-based UAN to perform efficient underwater data collection, reduce data collection time and energy consumption, and balance data collection tasks in the AUV-based UAN.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:14 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617907</guid>
    </item>
    <item>
      <title>A Lightweight Transformer-Based Collision Detection and Load Estimation Scheme for Massive Random Access in 6G Satellite-Ground Integrated Vehicular Networks</title>
      <link>https://trid.trb.org/View/2617884</link>
      <description><![CDATA[As an indispensable component of the 6G-enabled intelligent transportation systems, the satellite-ground integrated vehicular networks (SGIVN) have attracted widespread attention in recent years for its ability to provide continuous and ubiquitous connectivity services. However, in view of a huge number of access requirements from vehicle terminals and the restricted contention resources, the conventional random access (RA) schemes will suffer from severe overload issues when applied to the emerging SGIVN. To address this challenge, we propose a novel deep learning (DL) assisted collision detection and load estimation scheme to efficiently support massive access in the SGIVN. Specifically, a reliable RA preamble based on cyclically shifted Zadoff-Chu sequences is first designed as the precondition of collision detection, which can achieve an optimal performance trade-off between interference mitigation and user identification. By making full use of the intrinsic properties of preamble correlation results and the relevance analysis capability of attention mechanism, we further present a correlation feature extraction based deep RA collision detection framework embedded with a lightweight transformer network, thereby enabling the global dependencies of the few and important features associated with collided loads to be thoroughly acquired from the local correlation results with low overhead. Extensive simulation results validate the feasibility of our scheme in high-dynamic non-terrestrial network scenarios involving large-scale RA collisions, and demonstrate that it can obtain remarkably enhanced detection performance with short computational time, in comparison with state-of-the-art DL-based schemes.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:13 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617884</guid>
    </item>
    <item>
      <title>Multimodal Localization Distillation Method for 3D Single Object Tracking Task in Intelligent Transportation Systems</title>
      <link>https://trid.trb.org/View/2617882</link>
      <description><![CDATA[In 3D intelligent transportation systems, effectively using multimodal data from cameras and LiDAR for object tracking is a key research area. Current methods struggle due to an overreliance on 3D LiDAR data that overlooks detailed 2D camera imagery and inefficient use of localization data, resulting in poor frame-by-frame object tracking. Therefore, we propose a Multimodal Localization Fusion Tracking (MLFT) method to overcome these problems in intelligent transportation systems. The proposed method enables precise localization of multimodal information inputs through a novel Multimodal Localization Distillation (MMLD) approach and achieves accurate frame-to-frame object tracking via an advanced Multimodal Knowledge Transform-Fusion Tracking (MKTF) technique. The proposed method is evaluated using three representative datasets: KITTI, NuScenes, and Waymo Open. The experimental results show that our proposed MLFT algorithm has strong performance, improving the success rate by about 3.0%, 4.1%, and 3.3% on the KITTI, NuScenes, and Waymo Open datasets, respectively, while improving the accuracy by about 3.2%, 3.1%, and 4.1%.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:13 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617882</guid>
    </item>
    <item>
      <title>Zone-YOLO: Vision-Language Object Detection Using Zone Prompt</title>
      <link>https://trid.trb.org/View/2617874</link>
      <description><![CDATA[Object detection in complex traffic scenarios is crucial for Intelligent Transportation Systems (ITS). At present, most real-time traffic object detection methods primarily rely on YOLO-style vision-only detectors, limiting their potential for further improvement. Vision-Language Object Detection (VLOD) has made promising progress currently, yet its adoption in the realm of ITS remains limited. Previous VLOD methods utilize text features in the classification task, without fully exploring their impact on the regression process for object localization. Besides, existing multi-modal fusion approaches fail to fuse text features with multi-scale image features at corresponding scales, which is detrimental to the representation capability of the model. In this work, we dive into the limitations above and introduce Zone-YOLO to improve the VLOD to a new level. Specifically, we propose Scale-Aware Modal Fusion (SAMF) to fully exploit the text and image features and learn to fuse the multi-modal representations seamlessly at different scales with channel- and modal-wise enhancement. Moreover, we present a novel Zone Prompt learning method to introduce text features into regression process and capture the zone-class-entity triple co-occurrence, which significantly improves the localization performance of the model. Extensive experiments show that Zone-YOLO outperforms the comparative methods by a considerable margin, achieving 55.1 AP, 72.1 AP50 and 71.2 APL on COCO. The competitive results on BDD100K and VisDrone2019 further demonstrate the superiority of Zone-YOLO on efficient traffic object detection.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:13 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617874</guid>
    </item>
    <item>
      <title>Attention Transfer-Based Deep Distilled Architecture for 6G Driven-Smart Vehicle Transportation System</title>
      <link>https://trid.trb.org/View/2617867</link>
      <description><![CDATA[Autonomous 6G-enabled Vehicle Transportation System (VTS) is receiving significant attention from researchers to ensure robust and safe driving operations. 6G-supported communication technologies show remarkable advancements in several vehicular domains, including automated and accurate identification of lanes, vehicles, traffic signs, and obstacles within the vehicle’s proximity. Integrating IoT and AI technologies can leverage the extensive information gathered by Autonomous Vehicles (AVs) for precise vehicle detection. Despite the rapid advancements in object detection for completely visible objects from oncoming vehicles, detecting objects in low-visibility environments remains a challenging task, especially at night. This paper presents an efficient vehicle road cooperation method using automated real-time roadside object identification system by processing the captured video frames from LiDAR sensors. The developed method leverages the strength of the teacher-student-based distilled deep learning architecture for precisely identifying roadside objects. The teacher model utilizes the improved weighting factor to lower the false positives for better feature refinement. Meanwhile, the student model is equipped with convolutional attention and hierarchical feature fusion to capture balanced positional and semantic discriminative and multi-scale feature maps. Further, a gradient-based attention transfer mechanism has been utilized for significant knowledge transfer using attention maps from the teacher-to-student model to capture spatial feature information for night vision. Extensive experimental results demonstrate that the developed method overshadows the state-of-the-art object detection methods by achieving 41.45%, and 55.44% mAP and mAR, respectively. Additionally, the efficacy of the developed method has been validated by incorporating the different use cases in 6G-enabled VTS.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:13 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617867</guid>
    </item>
    <item>
      <title>Multi-Source Patch Feature Fusion With Neighborhood Flash Attention Transformer for Pixel-Level Vehicle and Road Recognition in Hyperspectral Image</title>
      <link>https://trid.trb.org/View/2617865</link>
      <description><![CDATA[Hyperspectral imaging can capture the spectrum of each pixel in an image across various wavelengths, providing unparalleled opportunities for precise detection, classification, and analysis of transportation infrastructure. However, traditional methods often struggle with the curse of dimensionality, inter-class variability, and the spectral-spatial trade-off inherent in hyperspectral data. To address these challenges, we introduce a novel Multi-Source Patch Feature fusion based Neighborhood Flash Attention Transformer (MSPF-NFAT) for pixel-level vehicle and road recognition in hyperspectral images (HSIs). Our methodology hinges on the insight that the integration of complementary features from multiple sources and scales can significantly enhance classification performance. Specifically, the MSPF is designed to aggregate and harmonize features extracted from both spectral and spatial dimensions, as well as from different contextual scales within the image. This fusion process ensures a richer representation of the data, capturing both the fine-grained details and the broader contextual information essential for accurate classification. Building upon this enriched feature set, we employ the NFAT, a state-of-the-art attention mechanism that focuses on capturing local spatial relationships while efficiently scaling to accommodate the high-resolution characteristics of hyperspectral data. In addition, extensive experimental results on four widely used HSIs datasets show that our newly proposed method provides superior performance compared to other state-of-the-art methods.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:13 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617865</guid>
    </item>
    <item>
      <title>MARL-Based AUV Formation for Underwater Intelligent Autonomous Transport Systems Supported by 6G Network</title>
      <link>https://trid.trb.org/View/2617861</link>
      <description><![CDATA[With the advancement of communication technology from 5G to 6G, future communication networks will no longer be limited to land and air, and the ocean will also become the battlefield for 6G networks. The expansion of the network has expanded the scope of Intelligent Autonomous Transport Systems (IATS). As a new type of underwater transport system, Autonomous Underwater Vehicle (AUV) has gained popularity due to their advantages of autonomy, endurance, and concealment. In practical applications, it is necessary to fully consider the impact of uncertain marine environments on AUV’s motion, and also design stable control unit to achieve AUV formation. The core of the control unit is the AUV formation control algorithm, which should enable AUV to complete path planning and obstacle avoidance while ensuring formation control. In order to solve the above problems, an Intelligent Multi-agent path planning and formation control algorithm based on Value-decomposition networks (IMV) is proposed in this paper. Specifically, a three-dimensional high-resolution marine simulation environment located in the Mariana Trench is established, the state transition function and reward function are well designed under uncertain conditions for stable Multi-Agent Reinforcement Learning (MARL) mechanism, a Value-Decomposition Networks (VDN) based training framework is constructed to improve the convergence speed of the proposed method. The experimental results verify the excellent performance of the IMV method proposed in this paper, demonstrating that our method can outperform other methods in the aspect of stability, adaptability, intelligence, and timeliness.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:12 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617861</guid>
    </item>
    <item>
      <title>CFDHI-Net: Correlation-Driven Feature Decoupling and Hierarchical Integration Network for RGB-Thermal Semantic Segmentation</title>
      <link>https://trid.trb.org/View/2617847</link>
      <description><![CDATA[The fusion of RGB and thermal modal features for RGB-T semantic segmentation has been demonstrated to be an effective approach in complex road conditions and extreme lighting conditions. However, effectively integrating the distinct characteristics of different modals remains a significant challenge. Previous feature fusion techniques often overlook the correlations between multi-modal features, resulting in insufficient feature fusion. To tackle this challenge, this paper proposes a Correlation-driven Feature Decoupling and Hierarchical Integration Network term as CFDHI-Net. Initially, the multi-modal features are decoupled into strongly correlated and weakly correlated features through a Correlation-driven Multi-modal Feature Decoupling module. The Hierarchical Feature Integration layer is then proposed to fuse the correlation-drvien hierarchical features. Specifically, it incorporates both a Multi-modal Cosine Similarity Feature Fusion module for fusing strongly correlated features and a Multi-modal Differential Feature Fusion module for fusing weakly correlated features. Ultimately, a Hierarchical Feature Fusion module is designed in the layer to integrate the two fused hierarchical features, thereby achieving efficient multi-modal feature fusion. Furthermore, to reduce the model complexity of the network, we extend the VMamba backbone with linear complexity into a dual-branch structure, serving as the encoder of the CFDHI-Net. Experimental evaluations conducted on various multi-modal semantic segmentation datasets demonstrate that CFDHI-Net achieves the state-of-the-art performance in both segmentation accuracy and model efficiency. The code is available at https://github.com/donggaomu/CFDHI-Net]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:12 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617847</guid>
    </item>
    <item>
      <title>MPC-DS: A Safe Path Tracking Method for AGVs in Dynamic Environments With Dense Obstacles</title>
      <link>https://trid.trb.org/View/2617831</link>
      <description><![CDATA[In narrow environments with dynamic dense obstacles, it is difficult to find feasible paths for autonomous ground vehicles (AGVs) by using existing methods due to the strict security constraints on AGV movements. To overcome such difficulty, this work proposes an improved local path tracking algorithm based on model predictive control and a dynamic control barrier function with slack variable (MPC-DS). In this algorithm, slack variable is integrated with the control barrier function to convert the strict constraints to soft ones. To regulate the values of slack variable, a suitable penalty coefficient selected by using a series of comparative simulations is incorporated into MPC’s cost function. To test effectiveness of the proposed method, it is compared with three mainstream methods in environments with dense and sparse dynamic obstacles. Results of simulations and physical experiments show that AGV controlled by the proposed algorithm can avoid obstacles safely and efficiently in complex environments. It is worth noting that its use reduces energy consumption by 21.1% in comparison with the existing ones. To facilitate further research, our code is publicly available at https://github.com/BUCT-RobotLab417/MPC-DS]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:12 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617831</guid>
    </item>
    <item>
      <title>A Multiple Artificial Potential Functions Approach for Collision Avoidance in UAV Systems</title>
      <link>https://trid.trb.org/View/2617812</link>
      <description><![CDATA[Collision avoidance is a problem largely studied in robotics, particularly in uncrewed aerial vehicle (UAV) applications. The main challenges in this area are hardware limitations, the need for rapid response, and the uncertainty associated with obstacle detection. Artificial potential functions (APOFs) are a prominent method to address these challenges. However, existing solutions lack assurances regarding closed-loop stability and may result in chattering effects. Hence, we propose a high-level control method for static obstacle avoidance based on multiple artificial potential functions (MAPOFs), with a set of switching rules with conditions on the parameter tuning ensuring the stability of the final position. The stability proof is established by analyzing the closed-loop system using tools from hybrid systems theory. Furthermore, we validate the performance of the MAPOF control through simulations and real-life experiments, showcasing its effectiveness in avoiding static obstacles.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:11 GMT</pubDate>
      <guid>https://trid.trb.org/View/2617812</guid>
    </item>
    <item>
      <title>AUTOMODAL - Automation of Inland Container Terminal Cranes Solving People Detection in the Operating Range</title>
      <link>https://trid.trb.org/View/2671804</link>
      <description><![CDATA[The AutoModal project results show the prototypical end-to-end automation of the transshipment terminal. An essential component is the automation of a gantry crane, which was converted for this purpose so that independent automated processes could be carried out. The focus of the work was on the conversion and prototypical operation of the crane automation in a reference terminal. For this purpose, the gantry crane was equipped with additional sensors to ensure safe and reliable detection of persons in the crane environment and to enable automated operation. Suitable solutions were evaluated from a wide selection of sensors and tested on a test model. In parallel, suitable control software was developed, interfaces harmonized and hardware components integrated for operation. The software modules developed was published onto the GitLab Open Source platform. The following article describes the main outcome of the project. AutoModal was carried out as part of the funding programme for innovative port technologies of the Federal Ministry for Digital and Transport Affairs.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:10 GMT</pubDate>
      <guid>https://trid.trb.org/View/2671804</guid>
    </item>
    <item>
      <title>Dynamic 2D Scene Analysis Inside Train Compartments Using Deep Learning Techniques</title>
      <link>https://trid.trb.org/View/2671020</link>
      <description><![CDATA[This paper presents a comprehensive study on scene analysis inside a train using computer vision and deep learning techniques. The primary objective is to develop a dynamic 2D representation of the perceived scene by continuously projecting passengers onto a 2D plan. The research contributions focus on detecting abnormal situations inside a train, such as passenger discomfort and forgotten luggage. To achieve optimal results, we rigorously tested cutting-edge deep learning techniques, including convolutional neural networks (CNN) for object detection and tracking. Through fine-tuning pretrained models on annotated datasets, our system achieves precise localization of individuals, enabling accurate estimation of their actions. Additionally, temporal analysis is used to moderate anomaly detection mechanisms for identifying situations requiring attention. The AI-based system has undergone rigorous real-time evaluation using RGB cameras installed in an actual train car. Through an onboard train sever, our system proactively identifies potential abnormal situations and sends alert to a supervision application. During testing sessions on the train, several scenarios involving unattended luggage and passenger discomfort were successfully identified in real-time, with high confidence rates.]]></description>
      <pubDate>Tue, 30 Jun 2026 09:45:09 GMT</pubDate>
      <guid>https://trid.trb.org/View/2671020</guid>
    </item>
  </channel>
</rss>