Search

In the context of autonomous driving, real-time and accurate object detection is crucial for ensuring vehicle safety. To address the problem of severe missed detection and poor accuracy for small targets in complex traffic conditions, a novel object detection algorithm for automatic driving is proposed, named YED-YOLO. Firstly, based on the original algorithm, the efficient multiscale attention (EMA) module is introduced to enhance the attention to the underlying features. Then, the CSPDarknet53 to 2-stage FPN (C2f) module is improved to solve the problem of insufficient receptive fields by adding deformable convolutions (DCNs). Furthermore, a novel intersection over union (IoU) loss function calculation method is proposed, which improves the generalization ability and convergence speed of the model. Finally, the ablation and comparative experiments are conducted using the recall, precision and frames per second (FPS). The experimental results indicate that the algorithm proposed in this paper demonstrates substantial enhancement effet for smaller traffic participants such as pedestrians and cyclists, whose APs increase by 3% and 3.1%, respectively. [ABSTRACT FROM AUTHOR]

To address the high computational demands and slow convergence faced by DRL in the field of autonomous driving, this paper integrated VAE with PPO algorithm. By adopting VAE's feature encoding technology, it effectively transformed semantic images obtained from the Carla simulator into state inputs, thus tackling the high computational load of DRL in handling complex autonomous driving tasks. To solve the issues of local optima and slow convergence in DRL training, it introduced a driving intervention mechanism and a driver-guided experience replay mechanism. These mechanisms applied driving interventions during the initial training phase and when the model encounters local optima, so as to enhance the model's learning efficiency and generalization capability. Experimental validation, conducted in left-turn scenarios at intersections, shows that with the aid of the driving intervention mechanism, the model's performance improves more rapidly in the initial training phase. Moreover, driving interventions when encountering local optima further enhance the model's performance, with even more significant improvements observed in complex scenarios. [ABSTRACT FROM AUTHOR]

Correct and robust ego-lane index estimation is crucial for autonomous driving in the absence of high-definition maps, especially in urban environments. Previous ego-lane index estimation approaches rely on feature extraction, which limits the robustness. To overcome these shortages, this study proposes a robust ego-lane index estimation framework upon only the original visual image. After optimization of the processing route, the raw image was randomly cropped in the height direction and then input into a double supervised LaneLoc network to obtain the index estimations and confidences. A post-process was also proposed to achieve the global ego-lane index from the estimated left and right indexes with the total lane number. To evaluate our proposed method, we manually annotated the ego-lane index of public datasets which can work as an ego-lane index estimation baseline for the first time. The proposed algorithm achieved 96.48/95.40% (precision/recall) on the CULane dataset and 99.45/99.49% (precision/recall) on the TuSimple dataset, demonstrating the effectiveness and efficiency of lane localization in diverse driving environments. The code and dataset annotation results will be exposed publicly on . [ABSTRACT FROM AUTHOR]

Ground segmentation is crucial for guiding mobile robots and identifying nearby objects. However, it should be noted that the ground often presents complex topographical features, such as slopes and rugged terrains, which significantly increase the challenges associated with accurate ground segmentation tasks. To address this issue, we propose a novel approach to achieve rapid ground segmentation. The proposed method uses a multi-partition approach to extract ground points for each partition, followed by assessing the correction plane based on geometric characteristics of the ground surface and similarity among adjacent planes. An adaptive threshold is also introduced to enhance efficiency in extracting complex urban pavement. Our method was benchmarked against several contemporary techniques on the SemanticKITTI dataset. The precision was elevated by 1.72 % , and the precision deviation was diminished by 1.02 % , culminating in the most accurate and robust outcomes among the evaluated methods. [ABSTRACT FROM AUTHOR]

Objective: Varying driver distraction algorithms were developed using vehicle kinematics and driver gaze data obtained from a camera-based driver monitoring system (DMS). Background: Distracted driving characteristics can be difficult to accurately detect due to wide variation in driver behavior across driving environments. The growing availability of information about drivers and their involvement in the driving task increases the opportunity for accurately recognizing attention state. Method: A baseline for driver distraction levels was developed using a video feed of 24 separate drivers in varying naturalistic driving conditions. This initial assessment was used to develop four buffer-based algorithms that aimed to determine a driver's real-time attentiveness, via a variety of metrics and combinations thereof. Results: Of those tested, the optimal algorithm included ungrouped glance locations and speed. Notably, as an algorithm's performance of detecting very distracted drivers improved, its accuracy for correctly identifying attentive drivers decreased. Conclusion: At a minimum, drivers' gaze position and vehicle speed should be included when designing driver distraction algorithms to delineate between glance patterns observed at high and low speeds. Distraction algorithms should be designed with an understanding of their limitations, including instances in which they may fail to detect distracted drivers, or falsely notify attentive drivers. Application: This research adds to the body of knowledge related to driver distraction and contributes to available methods to potentially address and reduce occurrences. Machine learning algorithms can build on the data elements discussed to increase distraction detection accuracy using robust artificial intelligence. [ABSTRACT FROM AUTHOR]

Objective: Using brain haemodynamic responses to measure perceived risk from traffic complexity during automated driving. Background: Although well-established during manual driving, the effects of driver risk perception during automated driving remain unknown. The use of fNIRS in this paper for assessing drivers' states posits it could become a novel method for measuring risk perception. Methods: Twenty-three volunteers participated in an empirical driving simulator experiment with automated driving capability. Driving conditions involved suburban and urban scenarios with varying levels of traffic complexity, culminating in an unexpected hazardous event. Perceived risk was measured via fNIRS within the prefrontal cortical haemoglobin oxygenation and from self-reports. Results: Prefrontal cortical haemoglobin oxygenation levels significantly increased, following self-reported perceived risk and traffic complexity, particularly during the hazardous scenario. Conclusion: This paper has demonstrated that fNIRS is a valuable research tool for measuring variations in perceived risk from traffic complexity during highly automated driving. Even though the responsibility over the driving task is delegated to the automated system and dispositional trust is high, drivers perceive moderate risk when traffic complexity builds up gradually, reflected in a corresponding significant increase in blood oxygenation levels, with both subjective (self-reports) and objective (fNIRS) increasing further during the hazardous scenario. Application: Little is known regarding the effects of drivers' risk perception with automated driving. Building upon our experimental findings, future work can use fNIRS to investigate the mental processes for risk assessment and the effects of perceived risk on driving behaviours to promote the safe adoption of automated driving technology. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

To solve to the problems of long planning times and low success rates in automatic parking in constrained parking lanes, a modified approach to path planning was proposed, which could improve the hybrid A* algorithm. Firstly, the parking path was divided into two parts : a forward pose adjustment segment and a reverse parking segment. A linear-arc constrained optimization model was established to plan the pose adjustment segment, while finding a suitable starting point for reverse parking. Subsequently, collision risk cost was introduced as an additional factor in the hybrid A* algorithm. The node expansion process was improved by checking whether the vehicle contour intersects with obstacle lines, enhancing the real-time performance and safety of the reverse parking segment. Finally, a cost function was designed with criteria such as path length, smoothness, and deviation, taking into account the vehicle's kinematic constraints. A quadratic programming approach was used to smooth the initial path, resulting in the final path. Simulation analysis was conducted using MATLAB to compare the modified algorithm with the original algorithm. The results showed that, in constrained parking lanes, the improved algorithm could produce a smooth, collision-free parking path with a reduced search time of 23.8% compared with the hybrid A* algorithm. Additionally, the obtained path was safer and more suitable for tracking control. [ABSTRACT FROM AUTHOR]

Recent studies affirm the potential of autonomous vehicles (AVs) in reducing traffic accidents and fatalities. This article presents an overview of on-road AV testing and analyzes crash data involving AVs. The study emphasizes fatalities and errors, comparing AVs to conventional human-driven vehicles or nonautonomous vehicles (non-AVs). Statistical analysis indicates that in most cases, human errors predominantly cause accidents, with responsibility falling on non-AV entities, such as bicyclists and motorcyclists, as well as adverse weather and lighting conditions. Notably, rear-end collisions are prevalent. AVs display superior sensing and processing capabilities, except in adverse weather and automation failure scenarios. Surprisingly, human drivers are accountable for most accidents, overshadowing the significance of AV error rates. Common collision causes between AVs and non-AVs encompass overspeeding, inadequate following distances, and decision-making deficiencies in non-AVs. The study shows the potential of AVs to enhance road safety while shedding light on areas requiring continued improvement. [ABSTRACT FROM AUTHOR]

Automatic 3D object detection using monocular cameras presents significant challenges in the context of autonomous driving. Precise labeling of 3D object scales requires accurate spatial information, which is difficult to obtain from a single image due to the inherent lack of depth information in monocular images, compared to LiDAR data. In this paper, we propose a novel approach to address this issue by enhancing deep neural networks with depth information for monocular 3D object detection. The proposed method comprises three key components: 1)Feature Enhancement Pyramid Module: We extend the conventional Feature Pyramid Networks (FPN) by introducing a feature enhancement pyramid network. This module fuses feature maps from the original pyramid and captures contextual correlations across multiple scales. To increase the connectivity between low-level and high-level features, additional pathways are incorporated. 2)Auxiliary Dense Depth Estimator: We introduce an auxiliary dense depth estimator that generates dense depth maps to enhance the spatial perception capabilities of the deep network model without adding computational burden. 3)Augmented Center Depth Regression: To aid center depth estimation, we employ additional bounding box vertex depth regression based on geometry. Our experimental results demonstrate the superiority of the proposed technique over existing competitive methods reported in the literature. The approach showcases remarkable performance improvements in monocular 3D object detection, making it a promising solution for autonomous driving applications. [ABSTRACT FROM AUTHOR]

With advancements in autonomous driving technology, the variety of activities that can be performed in a vehicle has increased. This improves the possibility of watching virtual reality (VR) content on a head-mounted display (HMD). However, unlike VR used in stationary environments, in-car VR can lead to discomfort and motion sickness due to the vehicle movements. Additionally, the obstruction of the outside view during driving may cause user anxiety. In this study, we investigated, for the first time, the effect of dynamic road environments, such as turns, stops, and speed bumps, on the in-car VR experience. Based on our findings, we included situational awareness (SA) cues in the in-car VR content to help users perceive their surroundings and improve the user experience. We conducted a user study with thirty participants to validate the impact of these cues. Consequently, we discovered that the Dynamics cue, which provides SA information while maintaining the context of the VR content, improves user immersion and trust while easing VR motion sickness. [ABSTRACT FROM AUTHOR]

With the rapid promotion of autonomous driving technology, it is extremely important to scientifically anticipate the related technologies and analyze their possible impact on urban road systems. The accuracy of detection and localization of traffic elements of autonomous driving is closely related to the ability of autonomous driving devices to make control decisions and the safety of autonomous driving. The study designs a new high-speed road driving scheme based on autonomous driving by analyzing the challenges related to urban traffic that may be brought about by unmanned driving. On the basis of the faster R-CNN algorithm, the context information around the target is introduced to locate and detect small-scale traffic signs. A new pedestrian detection model is designed, which is based on the feature pyramid network and introduces the SE module to highlight the features of the visible part of the pedestrian and reduce the missed detection rate caused by inter-class occlusion. The improved traffic sign detection framework improves the detection accuracy by 18.91% compared to the original faster R-CNN, while the enhanced pedestrian inspection method improves the detection accuracy by 14.00%. For both traffic sign detection and pedestrian detection accuracy and speed are improved compared to the original method. [ABSTRACT FROM AUTHOR]

3D object-detection based on LiDAR point clouds can help driverless vehicles detect obstacles. However, the existing point-cloud-based object-detection methods are generally ineffective in detecting small objects such as pedestrians and cyclists. Therefore, a small-object-detection algorithm based on clustering is proposed. Firstly, a new segmented ground-point clouds segmentation algorithm is proposed, which filters out the object point clouds according to the heuristic rules and realizes the ground segmentation by multi-region plane-fitting. Then, the small-object point cloud is clustered using an improved DBSCAN clustering algorithm. The K-means++ algorithm for pre-clustering is used, the neighborhood radius is adaptively adjusted according to the distance, and the core point search method of the original algorithm is improved. Finally, the detection of small objects is completed using the directional wraparound box model. After extensive experiments, it was shown that the precision and recall of our proposed ground-segmentation algorithm reached 91.86% and 92.70%, respectively, and the improved DBSCAN clustering algorithm improved the recall of pedestrians and cyclists by 15.89% and 9.50%, respectively. In addition, visualization experiments confirmed that our proposed small-object-detection algorithm based on the point-cloud clustering method can realize the accurate detection of small objects. [ABSTRACT FROM AUTHOR]

In infrared detection scenarios, detecting and recognizing low-contrast and small-sized targets has always been a challenge in the field of computer vision, particularly in complex road traffic environments. Traditional target detection methods usually perform poorly when processing infrared small targets, mainly due to their inability to effectively extract key features and the significant feature loss that occurs during feature transmission. To address these issues, this paper proposes a fast detection and recognition model based on a multi-scale self-attention mechanism, specifically for small road targets in infrared detection scenarios. We first introduce and improve the DyHead structure based on the YOLOv8 algorithm, which employs a multi-head self-attention mechanism to capture target features at various scales and enhance the model's perception of small targets. Additionally, to prevent information loss during the feature transmission process via the FPN structure in traditional YOLO algorithms, this paper introduces and enhances the Gather-and-Distribute Mechanism. By computing dependencies between features using self-attention, it reallocates attention weights in the feature maps to highlight important features and suppress irrelevant information. These improvements significantly enhance the model's capability to detect small targets. Moreover, to further increase detection speed, we pruned the network architecture to reduce computational complexity and parameter count, making the model suitable for real-time processing scenarios. Experiments on our self built infrared road traffic dataset (mainly including two types of targets: vehicles and people) show that compared with the baseline, our method achieves a 3.1% improvement in AP and a 2.5% increase in mAP on the VisDrone2019 dataset, showing significant enhancements in both detection accuracy and processing speed for small targets, with improved robustness and adaptability. [ABSTRACT FROM AUTHOR]

To achieve Level 4 and above autonomous driving, a robust and stable autonomous driving system is essential to adapt to various environmental changes. This paper aims to perform vehicle pose estimation, a crucial element in forming autonomous driving systems, more universally and robustly. The prevalent method for vehicle pose estimation in autonomous driving systems relies on Real-Time Kinematic (RTK) sensor data, ensuring accurate location acquisition. However, due to the characteristics of RTK sensors, precise positioning is challenging or impossible in indoor spaces or areas with signal interference, leading to inaccurate pose estimation and hindering autonomous driving in such scenarios. This paper proposes a method to overcome these challenges by leveraging objects registered in a high-precision map. The proposed approach involves creating a semantic high-definition (HD) map with added objects, forming object-centric features, recognizing locations using these features, and accurately estimating the vehicle's pose from the recognized location. This proposed method enhances the precision of vehicle pose estimation in environments where acquiring RTK sensor data is challenging, enabling more robust and stable autonomous driving. The paper demonstrates the proposed method's effectiveness through simulation and real-world experiments, showcasing its capability for more precise pose estimation. [ABSTRACT FROM AUTHOR]

In the context of autonomous driving, the augmentation of existing data through simulations provides an elegant solution to the challenge of capturing the full range of adverse weather conditions in training datasets. However, existing physics-based augmentation models typically rely on single scattering approximations to predict light propagation under unfavorable conditions, such as fog. This can prevent the reproduction of important signal characteristics encountered in a real-world environment. Consequently, in this work, Monte Carlo simulations are employed to assess the relevance of multiple-scattered light to the detected LiDAR signal in different types of fog, with scattering phase functions calculated from Mie theory considering real particle size distributions. Bidirectional path tracing is used within the self-developed GPU-accelerated Monte Carlo software to compensate for the unfavorable photon statistics associated with the limited detection aperture of the LiDAR geometry. To validate the Monte Carlo software, an analytical solution of the radiative transfer equation for the time-resolved radiance in terms of scattering orders is derived, thereby providing an explicit representation of the double-scattered contributions. The results of the simulations demonstrate that the shape of the detected signal can be significantly impacted by multiple-scattered light, depending on LiDAR geometry and visibility. In particular, double-scattered light can dominate the overall signal at low visibilities. This indicates that considering higher scattering orders is essential for improving AI-based perception models. [ABSTRACT FROM AUTHOR]

This article introduces an open-source software stack designed for autonomous 1:10 scale model vehicles. Initially developed for the Bosch Future Mobility Challenge (BFMC) student competition, this versatile software stack is applicable to a variety of autonomous driving competitions. The stack comprises perception, planning, and control modules, each essential for precise and reliable scene understanding in complex environments such as a miniature smart city in the context of BFMC. Given the limited computing power of model vehicles and the necessity for low-latency real-time applications, the stack is implemented in C++, employs YOLO Version 5 s for environmental perception, and leverages the state-of-the-art Robot Operating System (ROS) for inter-process communication. We believe that this article and the accompanying open-source software will be a valuable resource for future teams participating in autonomous driving student competitions. Our work can serve as a foundational tool for novice teams and a reference for more experienced participants. The code and data are publicly available on GitHub. [ABSTRACT FROM AUTHOR]

Featured Application: The gained knowledge can be applied by road marking material manufacturers and automotive proving ground operators to provide road marking scenarios on test tracks to test perception systems that work with camera or LiDAR sensors. The investigated method is especially useful for urban scenarios with road markings that have complex geometric shapes, such as small radii, symbols, lettering, and specific corner shapes for parking slots. In order to test camera- and LiDAR-based perception of road markings for automated driving, vehicle developers have started to utilize concepts for the agile alteration of road marking patterns on proving grounds. Road marking materials commonly used within this concept are different types of tape that can easily be applied and removed on asphalt and concrete. Due to the elasticity of tape, it cannot be used efficiently for small radii, symbols, lettering, and specific corner shapes (e.g., for parking slots). These road marking patterns are common in urban environments. With the growing capability of automated driving systems and more applications for urban environments, edgy road marking shapes gain importance for proving ground testing. This study examines the use of water-soluble road marking paint specifically designed for the use case of temporary applications on proving grounds for camera- and LiDAR-based perception testing. We found that white, water-soluble paint with 1.5% binder content and 2.25% coalescing agent content can provide realistic road markings for vehicle testing purposes. However, solubility affects the paint's vulnerability to fading during rain. Hence, renewal activities over the course of longer test drives might be necessary. The paint could be removed using water pressure without significant residue or damaging of the asphalt. [ABSTRACT FROM AUTHOR]

Autonomous driving is a complex problem that requires intelligent decision making, and it has recently garnered significant interest due to its potential advantages in convenience and safety. In autonomous driving, conventional path planning to reach a destination is a time-consuming challenge. Therefore, learning-based approaches have been successfully applied to the controller or decision-making aspects of autonomous driving. However, these methods often lack explainability, as passengers cannot discern where the vehicle is headed. Additionally, most experiments primarily focus on highway scenarios, which do not effectively represent road curvature. To address these issues, we propose a reinforcement-learning-based trajectory learning in the Frenet frame (RLTF), which involves learning trajectories in the Frenet frame. Learning trajectories enable the consideration of future states and enhance explainability. We demonstrate that RLTF achieves a 100% success rate in the simulation environment, considering future states on curvy roads with continuous obstacles while overcoming issues associated with the Frenet frame. [ABSTRACT FROM AUTHOR]

Pedestrian trajectory prediction is crucial for developing collision avoidance algorithms in autonomous driving systems, aiming to predict the future movement of the detected pedestrians based on their past trajectories. The traditional methods for pedestrian trajectory prediction involve a sequence of tasks, including detection and tracking to gather the historical movement of the observed pedestrians. Consequently, the accuracy of trajectory prediction heavily relies on the accuracy of the detection and tracking models, making it susceptible to their performance. The prior research in trajectory prediction has mainly assessed the model performance using public datasets, which often overlook the errors originating from detection and tracking models. This oversight fails to capture the real-world scenario of inevitable detection and tracking inaccuracies. In this study, we investigate the cumulative effect of errors within integrated detection, tracking, and trajectory prediction pipelines. Through empirical analysis, we examine the errors introduced at each stage of the pipeline and assess their collective impact on the trajectory prediction accuracy. We evaluate these models across various custom datasets collected in Taiwan to provide a comprehensive assessment. Our analysis of the results derived from these integrated pipelines illuminates the significant influence of detection and tracking errors on downstream tasks, such as trajectory prediction and distance estimation. [ABSTRACT FROM AUTHOR]

The ability to make informed decisions in complex scenarios is crucial for intelligent automotive systems. Traditional expert rules and other methods often fall short in complex contexts. Recently, reinforcement learning has garnered significant attention due to its superior decision-making capabilities. However, there exists the phenomenon of inaccurate target network estimation, which limits its decision-making ability in complex scenarios. This paper mainly focuses on the study of the underestimation phenomenon, and proposes an end-to-end autonomous driving decision-making method based on an improved TD3 algorithm. This method employs a forward camera to capture data. By introducing a new critic network to form a triple-critic structure and combining it with the target maximization operation, the underestimation problem in the TD3 algorithm is solved. Subsequently, the multi-timestep averaging method is used to address the policy instability caused by the new single critic. In addition, this paper uses Carla platform to construct multi-vehicle unprotected left turn and congested lane-center driving scenarios and verifies the algorithm. The results demonstrate that our method surpasses baseline DDPG and TD3 algorithms in aspects such as convergence speed, estimation accuracy, and policy stability. [ABSTRACT FROM AUTHOR]

The rapid development of active safety systems in the automotive industry and research in autonomous driving requires reliable, high-precision sensors that provide rich information about the surrounding environment and the behaviour of other road users. In practice, there is always some non-zero mounting misalignment, i.e., angular inaccuracy in a sensor's mounting on a vehicle. It is essential to accurately estimate and compensate for this misalignment further programmatically (in software). In the case of radars, imprecise mounting may result in incorrect/inaccurate target information, problems with the tracking algorithm, or a decrease in the power reflected from the target. Sensor misalignment should be mitigated in two ways: through the correction of an inaccurate alignment angle via the estimated value of the misalignment angle or alerting other components of the system of potential sensor degradation if the misalignment is beyond the operational range. This work analyses misalignment's influences on radar sensors and other system components. In the mathematically proven example of a vertically misaligned radar, pedestrian detectability dropped to one-third of the maximum range. In addition, mathematically derived heading estimation errors demonstrate the impact on data association in data fusion. The simulation results presented show that the angle of misalignment exponentially increases the risk of false track splitting. Additionally, the paper presents a comprehensive review of radar alignment techniques, mostly found in the patent literature, and implements a baseline algorithm, along with suggested key performance indicators (KPIs) to facilitate comparisons for other researchers. [ABSTRACT FROM AUTHOR]

The automotive industry, with particular reference to the off-road sector, is facing several challenges, including the integration of Advanced Driver Assistance Systems (ADASs), the introduction of autonomous driving capabilities, and system-specific requirements that are different from the traditional car market. Current vehicular electrical–electronic (E/E) architectures are unable to support the amount of data for new vehicle functionalities, requiring the transition to zonal architectures, new communication standards, and the adoption of Drive-by-Wire technologies. In this work, we propose an automated methodology for next-generation off-road vehicle E/E architectural design. Starting from the regulatory requirements, we use a MILP-based optimizer to find candidate solutions, a discrete event simulator to validate their feasibility, and an ascent-based gradient method to reformulate the constraints for the optimizer in order to converge to the final architectural solution. We evaluate the results in terms of latency, jitter, and network load, as well as provide a Pareto analysis that includes power consumption, cost, and system weight. [ABSTRACT FROM AUTHOR]

Robust object detection in complex environments, poor visual conditions, and open scenarios presents significant technical challenges in autonomous driving. These challenges necessitate the development of advanced fusion methods for millimeter-wave (mmWave) radar point cloud data and visual images. To address these issues, this paper proposes a radar–camera robust fusion network (RCRFNet), which leverages self-supervised learning and open-set recognition to effectively utilise the complementary information from both sensors. Specifically, the network uses matched radar–camera data through a frustum association approach to generate self-supervised signals, enhancing network training. The integration of global and local depth consistencies between radar point clouds and visual images, along with image features, helps construct object class confidence levels for detecting unknown targets. Additionally, these techniques are combined with a multi-layer feature extraction backbone and a multimodal feature detection head to achieve robust object detection. Experiments on the nuScenes public dataset demonstrate that RCRFNet outperforms state-of-the-art (SOTA) methods, particularly in conditions of low visual visibility and when detecting unknown class objects. [ABSTRACT FROM AUTHOR]

Despite impressive progress, obtaining appropriate data for instance-level lane segmentation remains a significant challenge. This limitation hinders the refinement of granular lane-related applications such as lane line crossing surveillance, pavement maintenance, and management. To address this gap, we introduce a benchmark for lane instance segmentation called InstLane. To the best of our knowledge, InstLane constitutes the first publicly accessible instance-level segmentation standard for lane line detection. The complexity of InstLane emanates from the fact that the original data are procured using cameras mounted laterally, as opposed to traditional front-mounted sensors. InstLane encapsulates a range of challenging scenarios, enhancing the generalization and robustness of the lane line instance segmentation algorithms. In addition, we propose GeoLaneNet, a real-time, geometry-aware lane instance segmentation network. Within GeoLaneNet, we design a finer localization of lane proto-instances based on geometric features to counteract the prevalent omission or multiple detections in dense lane scenarios resulting from non-maximum suppression (NMS). Furthermore, we present a scheme that employs a larger receptive field to achieve profound perceptual lane structural learning, thereby improving detection accuracy. We introduce an architecture based on partial feature transformation to expedite the detection process. Comprehensive experiments on InstLane demonstrate that GeoLaneNet can achieve up to twice the speed of current State-Of-The-Artmethods, reaching 139 FPS on an RTX3090 and a mask AP of 73.55%, with a permissible trade-off in AP, while maintaining comparable accuracy. These results underscore the effectiveness, robustness, and efficiency of GeoLaneNet in autonomous driving. [ABSTRACT FROM AUTHOR]

Objective: This study proposed a model to predict passenger motion sickness under the presence of a visual-vestibular conflict and assessed its performance with respect to previously recorded experimental data. Background: While several models have been shown useful to predict motion sickness under repetitive motion, improvements are still desired in terms of predicting motion sickness in realistic driving conditions. There remains a need for a model that considers angular and linear visual-vestibular motion inputs in three dimensions to improve prediction of passenger motion sickness. Method: The model combined the subjective vertical conflict theory and human motion perception models. The proposed model integrates visual and vestibular sensed 6 DoF motion signals in a novel architecture. Results: Model prediction results were compared to motion sickness data obtained from studies conducted in motion simulators as well as on-road vehicle testing, yielding trends that are congruent with observed results in both cases. Conclusion: The model demonstrated the ability to predict trends in motion sickness response for conditions in which a passenger performs a task on a handheld device versus facing forward looking ahead under realistic driving conditions. However, further analysis across a larger population is necessary to better assess the model's performance. Application: The proposed model can be used as a tool to predict motion sickness under different levels of visual-vestibular conflict. This can be leveraged to design interventions capable of mitigating passenger motion sickness. Further, this model can provide insights that aid in the development of passenger experiences inside autonomous vehicles. [ABSTRACT FROM AUTHOR]

• The level of urgency affects situation awareness(SA), trust, workload, and perceived urgency. • In high-level urgency, level 3 SA significantly improves situation awareness. • The displays with SA cues enhance system transparency, increasing drivers' trust. • Information that adequately describes the situation reduces the drivers' workload. • Level 1 SA is not significantly different from a display without SA factors. Semi-autonomous vehicles still require drivers to take over in an unexpected situation. In this situation, the increased cognitive load on the driver can lead to distraction, which in turn reduces situational awareness (SA) and prevents appropriate responses, increasing the risk of accidents. For this reason, providing interfaces that enhance SA is essential to ensuring safety and optimal performance. However, existing SA-based display research often overlooks the levels of SA and effectiveness of modalities, especially in urgent situations where a driver's SA might be compromised. The present study aims to design SA-based display that considers the urgency of the situation and the level of SA in line with a specific design framework and modality effectiveness. We conducted an experiment using simulated videos to evaluate the effectiveness of the SA-based display. This experiment assessed the effects of three urgent situations and three levels of situational awareness-based displays on drivers' SA, situational trust, mental workload, and perceived urgency. We employed a 3x4 mixed-factorial design for the experiment. The between-subject factors were the SA levels (perception, comprehension, and projection) and a baseline. The within-subject factors were urgency scenarios (low, medium, and high). The results showed that as urgency increased, displays reflecting Level 3 SA, which requires prediction, significantly improved SA compared to other displays. We expect our findings to contribute to the practical design of automotive displays by providing useful considerations for SA-based display design. [ABSTRACT FROM AUTHOR]

Deep learning approaches have been gaining importance in several applications. However, the widespread use of these methods in safety-critical domains, such as Autonomous Driving, is still dependent on their reliability and trustworthiness. The goal of this paper is to provide a review of deep learning-based uncertainty methods and their applications to support perception tasks for Autonomous Driving. We detail significant Uncertainty Quantification and calibration methods, and their contributions and limitations, as well as important metrics and concepts. We present an overview of the state of the art of out-of-distribution detection and active learning, where uncertainty estimates are commonly applied. We show how these methods have been applied in the automotive context, providing a comprehensive analysis of reliable AI for Autonomous Driving. Finally, challenges and opportunities for future work are discussed for each topic. [ABSTRACT FROM AUTHOR]

Copyright of Computer Measurement & Control is the property of Magazine Agency of Computer Measurement & Control and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Currently, many real-time semantic segmentation networks aim for heightened accuracy, inevitably leading to increased computational complexity and reduced inference speed. Therefore, striking a balance between accuracy and speed has emerged as a crucial concern in this domain. To address these challenges, this study proposes a dual-branch fusion network with multiscale atrous pyramid pooling aggregate contextual features for real-time semantic segmentation (MAFNet). The first key component, the semantics guide spatial-details module (SGSDM) not only facilitates precise boundary extraction and fine-grained classification, but also provides semantic-based feature representation, thereby enhancing support for spatial analysis and decision boundaries. The second component, the multiscale atrous pyramid pooling module (MSAPPM), is designed by combining dilation convolution with feature pyramid pooling operations at various dilation rates. This design not only expands the receptive field, but also aggregates rich contextual information more effectively. To further improve the fusion of feature information generated by the dual-branch, a bilateral fusion module (BFM) is introduced. This module employs cross-fusion by calculating weights generated by the dual-branch to balance the weight relationship between the dual branches, thereby achieving effective feature information fusion. To validate the effectiveness of the proposed network, experiments are conducted on a single A100 GPU. MAFNet achieves a mean intersection over union (mIoU) of 77.4% at 70.9 FPS on the Cityscapes test dataset and 77.6% mIoU at 192.5 FPS on the CamVid test dataset. The experimental results conclusively demonstrated that MAFNet effectively strikes a balance between accuracy and speed. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

A lightweight, high‐definition vector map (HDVM) enables fully autonomous vehicles. However, the generation of HDVM remains a challenging problem, especially in complex urban scenarios. Moreover, numerous factors in the urban environment can degrade the accuracy of HDVM, necessitating a reliable error quantification. To address these challenges, this paper presents an open‐source and generic HDVM generation pipeline that integrates the global navigation satellite system (GNSS), inertial navigation system (INS), light detection and ranging (LiDAR), and camera. The pipeline begins by extracting semantic information from raw images using the Swin Transformer. The absolute 3D information of semantic objects is then retrieved using depth from the 3D LiDAR, and pose estimation from GNSS/INS integrated navigation system. Vector information (VI), such as lane lines, is extracted from the semantic information to construct the HDVM. To assess the potential error of the HDVM, this paper systematically quantifies the impacts of two key error sources, segmentation and LiDAR‐camera extrinsic parameter error. An error propagation scheme is first formed to illustrate how these errors fundamentally influence the accuracy of the HDVM. The effectiveness of the proposed pipeline is demonstrated through our codeavailable at https://github.com/ebhrz/HDMap. The performance is verified using typical datasets, including indoor garages and complex urban scenarios. [ABSTRACT FROM AUTHOR]

Aiming to address the challenge where existing methods struggle to predict accurate disparities for imperfectly rectified stereo images, and that supervised training requires a considerable amount of ground truth, a self‐supervised binocular depth estimation algorithm with self‐rectification for autonomous driving is proposed. Firstly, a subnetwork dedicated to stereo rectification, aiming to estimate the homography between stereo images is developed. This homography facilitates the transformation of stereo image pairs, aligning their corresponding pixels horizontally. Secondly, a foundational self‐supervised framework primarily centred on minimizing errors in stereo image reconstruction, combined with the generative‐adversarial strategy is introduced. Finally, a vertical offset prediction module (VOPM) is incorporated into the basic framework to further enhance the resistance of the stereo matching network to pixel‐level vertical offset errors. Experimental results on the public KITTI dataset for autonomous driving demonstrate the effectiveness of this approach in improving the disparity prediction performance for imperfectly rectified stereo images. Moreover, the self‐supervised training framework exhibits superiority over state‐of‐the‐art methods. [ABSTRACT FROM AUTHOR]

To enhance the efficiency and accuracy of environmental perception for autonomous vehicles, we propose GDMNet, a unified multi-task perception network for autonomous driving, capable of performing drivable area segmentation, lane detection, and traffic object detection. Firstly, in the encoding stage, features are extracted, and Generalized Efficient Layer Aggregation Network (GELAN) is utilized to enhance feature extraction and gradient flow. Secondly, in the decoding stage, specialized detection heads are designed; the drivable area segmentation head employs DySample to expand feature maps, the lane detection head merges early-stage features and processes the output through the Focal Modulation Network (FMN). Lastly, the Minimum Point Distance IoU (MPDIoU) loss function is employed to compute the matching degree between traffic object detection boxes and predicted boxes, facilitating model training adjustments. Experimental results on the BDD100K dataset demonstrate that the proposed network achieves a drivable area segmentation mean intersection over union (mIoU) of 92.2%, lane detection accuracy and intersection over union (IoU) of 75.3% and 26.4%, respectively, and traffic object detection recall and mAP of 89.7% and 78.2%, respectively. The detection performance surpasses that of other single-task or multi-task algorithm models. [ABSTRACT FROM AUTHOR]

Copyright of Automotive Digest is the property of Automotive Digest Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The dramatic increase in car ownership, contrasted with the inadequate development of urban parking infrastructure, has led to significant parking challenges. Such challenges, including limited parking spaces and restricted parking areas, severely hinder the efficacy of vehicular travel. Automated parking, distinguished by its low velocity and advanced autonomy in certain environments, emerges as the forefront product in the realm of autonomous driving. The development status and trend of automatic parking market at home and abroad were analyzed, and the planning and decision module of parking system was mainly discussed. The importance of motion planning was emphasized, which directly affects the quality of vehicle driving. The research results at home and abroad were reviewed, including curve interpolation, random sampling, graph search, intelligent algorithm and numerical optimization method. Furthermore, the advantages and limitations of different path planning were analyzed. In conclusion, the future integration and development direction of automatic parking technology were prospected. [ABSTRACT FROM AUTHOR]

In complex urban scenarios like intersections without dedicated left-turn signals, the construction of planning systems that maximize efficiency while guarantee safety has been a significant challenge. In this paper, we propose a reinforcement learning approach based on curriculum learning using real world dataset, and we develop a partial end-to-end planning and control model capable of adapting to variable temporal and spatial dimensional state inputs, applying it to autonomous driving task. Our model is compared with mainstream reinforcement learning algorithms to validate that our proposed algorithm can effectively solve complex spatio-temporal planning problems. This significantly enhances the efficiency of passing while maintaining a certain level of safety. [ABSTRACT FROM AUTHOR]

With the increasing complexity of autonomous driving tasks, the computational demands on single vehicular computing units have escalated, more and more tasks need to be offloaded to the edge. These tasks vary in latency sensitivity: real-time tasks, critical for passenger safety, require strict deadline adherence, whereas the latency of standard tasks mainly affects the user experience and has more flexible constraints. Addressing the challenge of selecting suitable edge computing nodes to enhance the offloading success rate of real-time tasks amidst a vast and heterogeneous cluster becomes crucial. This paper introduces the adaptive priority-based hierarchical task offloading (APHTO) algorithm, which optimizes task offloading strategies by accounting for the diverse latency constraints of different task types. Experiments demonstrate that under optimal performance conditions, APHTO significantly outperforms existing algorithms such as Min–Min, Max–Min, CUS, and FMS in reducing task latency by 20.31%, increasing offloading success rates by 35.83%, and improving resource utilization by 30.21%, marking a substantial advancement in task offloading strategies for autonomous driving integrated with MEC. [ABSTRACT FROM AUTHOR]

Visual object tracking is an important technology in camera-based sensor networks, which has a wide range of practicability in auto-drive systems. A transformer is a deep learning model that adopts the mechanism of self-attention, and it differentially weights the significance of each part of the input data. It has been widely applied in the field of visual tracking. Unfortunately, the security of the transformer model is unclear. It causes such transformer-based applications to be exposed to security threats. In this work, the security of the transformer model was investigated with an important component of autonomous driving, i.e., visual tracking. Such deep-learning-based visual tracking is vulnerable to adversarial attacks, and thus, adversarial attacks were implemented as the security threats to conduct the investigation. First, adversarial examples were generated on top of video sequences to degrade the tracking performance, and the frame-by-frame temporal motion was taken into consideration when generating perturbations over the depicted tracking results. Then, the influence of perturbations on performance was sequentially investigated and analyzed. Finally, numerous experiments on OTB100, VOT2018, and GOT-10k data sets demonstrated that the executed adversarial examples were effective on the performance drops of the transformer-based visual tracking. White-box attacks showed the highest effectiveness, where the attack success rates exceeded 90% against transformer-based trackers. [ABSTRACT FROM AUTHOR]

Cognitive scientists believe that adaptable intelligent agents like humans perform spatial reasoning tasks by learned causal mental simulation. The problem of learning these simulations is called predictive world modeling. We present the first framework for a learning open-vocabulary predictive world model (OV-PWM) from sensor observations. The model is implemented through a hierarchical variational autoencoder (HVAE) capable of predicting diverse and accurate fully observed environments from accumulated partial observations. We show that the OV-PWM can model high-dimensional embedding maps of latent compositional embeddings representing sets of overlapping semantics inferable by sufficient similarity inference. The OV-PWM simplifies the prior two-stage closed-set PWM approach to the single-stage end-to-end learning method. CARLA simulator experiments show that the OV-PWM can learn compact latent representations and generate diverse and accurate worlds with fine details like road markings, achieving 69 mIoU over six query semantics on an urban evaluation sequence. We propose the OV-PWM as a versatile continual learning paradigm for providing spatio-semantic memory and learned internal simulation capabilities to future general-purpose mobile robots. [ABSTRACT FROM AUTHOR]

Recently, the growing demand for autonomous driving in the industry has led to a lot of interest in 3D object detection, resulting in many excellent 3D object detection algorithms. However, most 3D object detectors focus only on a single set of LiDAR points, ignoring their potential ability to improve performance by leveraging the information provided by the consecutive set of LIDAR points. In this paper, we propose a novel 3D object detection method called temporal motion-aware 3D object detection (TM3DOD), which utilizes temporal LiDAR data. In the proposed TM3DOD method, we aggregate LiDAR voxels over time and the current BEV features by generating motion features using consecutive BEV feature maps. First, we present the temporal voxel encoder (TVE), which generates voxel representations by capturing the temporal relationships among the point sets within a voxel. Next, we design a motion-aware feature aggregation network (MFANet), which aims to enhance the current BEV feature representation by quantifying the temporal variation between two consecutive BEV feature maps. By analyzing the differences and changes in the BEV feature maps over time, MFANet captures motion information and integrates it into the current feature representation, enabling more robust and accurate detection of 3D objects. Experimental evaluations on the nuScenes benchmark dataset demonstrate that the proposed TM3DOD method achieved significant improvements in 3D detection performance compared with the baseline methods. Additionally, our method achieved comparable performance to state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Accurate and comprehensive 3D objects detection is important for perception systems in autonomous driving. Nevertheless, contemporary mainstream methods tend to perform more effectively on large objects in regions proximate to the LiDAR, leaving limited exploration of long-range objects and small objects. The divergent point pattern of LiDAR, which results in a reduction in point density as the distance increases, leads to a non-uniform point distribution that is ill-suited to discretized volumetric feature extraction. To address this challenge, we propose the Foreground Voxel Proposal (FVP) module, which effectively locates and generates voxels at the foreground of objects. The outputs are subsequently merged to mitigating the difference in point cloud density and completing the object shape. Furthermore, the susceptibility of small objects to occlusion results in the loss of feature space. To overcome this, we propose the Multi-Scale Feature Integration Network (MsFIN), which captures contextual information at different ranges. Subsequently, the outputs of these features are integrated through a cascade framework based on transformers in order to supplement the object features space. The extensive experimental results demonstrate that our network achieves remarkable results. Remarkably, our approach demonstrated an improvement of 8.56% AP on the SECOND baseline for the Car detection task at a distance of more than 20 m, and 9.38% AP on the Cyclist detection task. [ABSTRACT FROM AUTHOR]

Autonomous driving involves collaborative data sensing and traffic sign recognition. Emerging artificial intelligence technology has brought tremendous advances to vehicular networks. However, it is challenging to guarantee privacy and security when using traditional centralized machine learning methods for traffic sign recognition. It is urgent to introduce a distributed machine learning approach to protect private data of connected vehicles. In this paper, we propose a local differential privacy-based binary encoding federated learning approach. The binary encoding techniques and random perturbation methods are used in distributed learning scenarios to enhance the efficiency and security of data transmission. For the vehicle layer in this approach, the model is trained locally, and the model parameters are uploaded to the central server through encoding and perturbing. The central server designs the corresponding decoding, correction scheme, and regression statistical method for the received binary string. Then, the model parameters are aggregated and updated in the server and transmitted to the vehicle until the learning model is trained. The performance of the proposed approach is verified using the German Traffic Sign Recognition Benchmark data set. The simulation results show that the convergence of the approach is better with the increase in the learning cycle. Compared with baseline methods, such as the convolutional neural network, random forest, and backpropagation, the proposed approach achieves higher accuracy in the process of traffic sign recognition, with an increase of 6%. [ABSTRACT FROM AUTHOR]

Aiming at the problem that the existing motion planning methods for autonomous driving ignore the interaction of surrounding vehicles when extracting state information and the bad planning effect in dense traffic scenarios, a motion planning model combined with graph neural network and deep reinforcement learning is proposed. Firstly, based on the graph neural network, an interactive feature representation method of self-driving vehicles is proposed to extract spatial interaction features of multiple traffic participants. In this case, a learning strategy for motion planning is designed based on twin delayed deep deterministic policy gradient (TD3), and the next action is predicted from the interactive features so as to realize motion planning. The proposed method is compared with the current motion planning model LSTM+TD3, TD3 and deep deterministic policy gradient (DDPG) for autonomous driving, in dense traffic scenarios, the experimental results of training and testing in the PGDrive driving simulator increased by 36%, 43%, 23% and 13, 19, 53 percentage points compared with the comparison method, which means the proposed method can effectively solve the problem of interactive information perception of surrounding vehicles for better motion planning of autonomous driving. [ABSTRACT FROM AUTHOR]

This paper introduces an anchorless deep learning model designed for efficient analysis and processing of large-scale 3D synthetic traffic sign board datasets. With an ever-increasing emphasis on autonomous driving systems and their reliance on precise environmental perception, the ability to accurately interpret traffic sign information is crucial. Our model seamlessly integrates object detection, depth estimation, deformable parts, and text character extraction functionalities, facilitating a comprehensive understanding of road signs in simulated environments that mimic the real world. The dataset used has a large number of artificially generated traffic signs for 183 different classes. The signs include place names in Japanese and English, expressway names in Japanese and English, distances and motorway numbers, and direction arrow marks with different lighting, occlusion, viewing angles, camera distortion, day and night cycles, and bad weather like rain, snow, and fog. This was done so that the model could be tested thoroughly in a wide range of difficult conditions. We developed a convolutional neural network with a modified lightweight hourglass backbone using depthwise spatial and pointwise convolutions, along with spatial and channel attention modules that produce resilient feature maps. We conducted experiments to benchmark our model against the baseline model, showing improved accuracy and efficiency in both depth estimation and text extraction tasks, crucial for real-time applications in autonomous navigation systems. With its model efficiency and partwise decoded predictions, along with Optical Character Recognition (OCR), our approach suggests its potential as a valuable tool for developers of Advanced Driver-Assistance Systems (ADAS), Autonomous Vehicle (AV) technologies, and transportation safety applications, ensuring reliable navigation solutions. [ABSTRACT FROM AUTHOR]

AbstractAutonomous driving (AD) technology has gradually matured, but the lack of trust and acceptance from users limits its adoption and diffusion. This paper aims to solve the challenges and alleviate the trust crisis by exploring reasonable interactive output modalities. Firstly, we propose a multimodal interactive output scheme with 9 different feedback level. Subsequently, we conduct simulated driving experiments on the scheme, using three methods: driving trust experience questionnaire (DTEQ), eye-tracking, and takeover desire recording. Finally, we analyze the trends and correlations of results at different interactive output levels. The results indicate that the user trust experience is highly positively correlated with the level of interactive output modalities. Under reasonable design, multimodal and high-level interactive output is beneficial for providing more comprehensive feedback information, reducing users’ visual workload, and improving trust. This work provides a foundation for enhancement of user trust experience, and helps promote the application and popularization of AD. [ABSTRACT FROM AUTHOR]

Accurate 3D image recognition, critical for autonomous driving safety, is shifting from the LIDAR-based point cloud to camera-based depth estimation technologies driven by cost considerations and the point cloud's limitations in detecting distant small objects. This research aims to enhance MDE (Monocular Depth Estimation) using a single camera, offering extreme cost-effectiveness in acquiring 3D environmental data. In particular, this paper focuses on novel data augmentation methods designed to enhance the accuracy of MDE. Our research addresses the challenge of limited MDE data quantities by proposing the use of synthetic-based augmentation techniques: Mask, Mask-Scale, and CutFlip. The implementation of these synthetic-based data augmentation strategies has demonstrably enhanced the accuracy of MDE models by 4.0% compared to the original dataset. Furthermore, this study introduces the RMS (Real-time Monocular Depth Estimation configuration considering Resolution, Efficiency, and Latency) algorithm, designed for the optimization of neural networks to augment the performance of contemporary monocular depth estimation technologies through a three-step process. Initially, it selects a model based on minimum latency and REL criteria, followed by refining the model's accuracy using various data augmentation techniques and loss functions. Finally, the refined model is compressed using quantization and pruning techniques to minimize its size for efficient on-device real-time applications. Experimental results from implementing the RMS algorithm indicated that, within the required latency and size constraints, the IEBins model exhibited the most accurate REL (absolute RELative error) performance, achieving a 0.0480 REL. Furthermore, the data augmentation combination of the original dataset with Flip, Mask, and CutFlip, alongside the SigLoss loss function, displayed the best REL performance, with a score of 0.0461 . The network compression technique using FP16 was analyzed as the most effective, reducing the model size by 83.4% compared to the original while maintaining the least impact on REL performance and latency. Finally, the performance of the RMS algorithm was validated on the on-device autonomous driving platform, NVIDIA Jetson AGX Orin, through which optimal deployment strategies were derived for various applications and scenarios requiring autonomous driving technologies. [ABSTRACT FROM AUTHOR]

Reinforcement Learning (RL) methods are regarded as effective for designing autonomous driving policies. However, even when RL policies are trained to convergence, ensuring their robust safety remains a challenge, particularly in long-tail data. Therefore, decision-making based on RL must adequately consider potential variations in data distribution. This paper presents a framework for highway autonomous driving decisions that prioritizes both safety and robustness. Utilizing the proposed Replay Buffer Constrained Policy Optimization (RECPO) method, this framework updates RL strategies to maximize rewards while ensuring that the policies always remain within safety constraints. We incorporate importance sampling techniques to collect and store data in a Replay buffer during agent operation, allowing the reutilization of data from old policies for training new policy models, thus mitigating potential catastrophic forgetting. Additionally, we transform the highway autonomous driving decision problem into a Constrained Markov Decision Process (CMDP) and apply our proposed RECPO for training, optimizing highway driving policies. Finally, we deploy our method in the CARLA simulation environment and compare its performance in typical highway scenarios against traditional CPO, current advanced strategies based on Deep Deterministic Policy Gradient (DDPG), and IDM + MOBIL (Intelligent Driver Model and the model for minimizing overall braking induced by lane changes). The results show that our framework significantly enhances model convergence speed, safety, and decision-making stability, achieving a zero-collision rate in highway autonomous driving. [ABSTRACT FROM AUTHOR]

Background: High-definition maps can provide necessary prior data for autonomous driving, as well as the corresponding beyond-line-of-sight perception, verification and positioning, dynamic planning, and decision control. It is a necessary element to achieve L4/L5 unmanned driving at the current stage. However, currently, high-definition maps still have problems such as a large amount of data, a lot of data redundancy, and weak data correlation, which make autonomous driving fall into difficulties such as high data query difficulty and low timeliness. In order to optimize the data quality of high-definition maps, enhance the degree of data correlation, and ensure that they better assist vehicles in safe driving and efficient passage in the autonomous driving scenario, it is necessary to clarify the information system thinking of high-definition maps, propose a complete and accurate model, determine the content and functions of each level of the model, and continuously improve the information system model. Objective: The study aimed to put forward a complete and accurate high-definition map information system model and elaborate in detail the content and functions of each component in the data logic structure of the system model. Methods: Through research methods such as the modeling method and literature research method, we studied the high-definition map information system model in the autonomous driving scenario and explored the key technologies therein. Results: We put forward a four-layer integrated high-definition map information system model, elaborated in detail the content and functions of each component (map, road, vehicle, and user) in the data logic structure of the model, and also elaborated on the mechanism of the combined information of each level of the model to provide services in perception, positioning, decision making, and control for autonomous driving vehicles. This article also discussed two key technologies that can support autonomous driving vehicles to complete path planning, navigation decision making, and vehicle control in different autonomous driving scenarios. Conclusions: The four-layer integrated high-definition map information model proposed by this research institute has certain application feasibility and can provide references for the standardized production of high-definition maps, the unification of information interaction relationships, and the standardization of map data associations. [ABSTRACT FROM AUTHOR]