Showing papers in "IEEE Transactions on Intelligent Transportation Systems in 2020"
TL;DR: In this article, a novel neural network-based traffic forecasting method, the temporal graph convolutional network (T-GCN) model, which is combined with the graph convolutionsal network and the gated recurrent unit (GRU), is proposed.
Abstract: Accurate and real-time traffic forecasting plays an important role in the intelligent traffic system and is of great significance for urban traffic planning, traffic management, and traffic control. However, traffic forecasting has always been considered an “open” scientific issue, owing to the constraints of urban road network topological structure and the law of dynamic change with time. To capture the spatial and temporal dependences simultaneously, we propose a novel neural network-based traffic forecasting method, the temporal graph convolutional network (T-GCN) model, which is combined with the graph convolutional network (GCN) and the gated recurrent unit (GRU). Specifically, the GCN is used to learn complex topological structures for capturing spatial dependence and the gated recurrent unit is used to learn dynamic changes of traffic data for capturing temporal dependence. Then, the T-GCN model is employed to traffic forecasting based on the urban road network. Experiments demonstrate that our T-GCN model can obtain the spatio-temporal correlation from traffic data and the predictions outperform state-of-art baselines on real-world traffic datasets. Our tensorflow implementation of the T-GCN is available at https://www.github.com/lehaifeng/T-GCN .
1,188 citations
TL;DR: A novel deep learning framework, Traffic Graph Convolutional Long Short-Term Memory Neural Network (TGC-LSTM), to learn the interactions between roadways in the traffic network and forecast the network-wide traffic state and shows that the proposed model outperforms baseline methods on two real-world traffic state datasets.
Abstract: Traffic forecasting is a particularly challenging application of spatiotemporal forecasting, due to the time-varying traffic patterns and the complicated spatial dependencies on road networks. To address this challenge, we learn the traffic network as a graph and propose a novel deep learning framework, Traffic Graph Convolutional Long Short-Term Memory Neural Network (TGC-LSTM), to learn the interactions between roadways in the traffic network and forecast the network-wide traffic state. We define the traffic graph convolution based on the physical network topology. The relationship between the proposed traffic graph convolution and the spectral graph convolution is also discussed. An L1-norm on graph convolution weights and an L2-norm on graph convolution features are added to the model’s loss function to enhance the interpretability of the proposed model. Experimental results show that the proposed model outperforms baseline methods on two real-world traffic state datasets. The visualization of the graph convolution weights indicates that the proposed framework can recognize the most influential road segments in real-world traffic networks.
611 citations
TL;DR: In this article, a fully scalable and decentralized MARL algorithm for the state-of-the-art deep RL agent, advantage actor critic (A2C), within the context of adaptive traffic signal control (ATSC) is presented.
Abstract: Reinforcement learning (RL) is a promising data-driven approach for adaptive traffic signal control (ATSC) in complex urban traffic networks, and deep neural networks further enhance its learning power. However, the centralized RL is infeasible for large-scale ATSC due to the extremely high dimension of the joint action space. The multi-agent RL (MARL) overcomes the scalability issue by distributing the global control to each local RL agent, but it introduces new challenges: now, the environment becomes partially observable from the viewpoint of each local agent due to limited communication among agents. Most existing studies in MARL focus on designing efficient communication and coordination among traditional Q-learning agents. This paper presents, for the first time, a fully scalable and decentralized MARL algorithm for the state-of-the-art deep RL agent, advantage actor critic (A2C), within the context of ATSC. In particular, two methods are proposed to stabilize the learning procedure, by improving the observability and reducing the learning difficulty of each local agent. The proposed multi-agent A2C is compared against independent A2C and independent Q-learning algorithms, in both a large synthetic traffic grid and a large real-world traffic network of Monaco city, under simulated peak-hour traffic dynamics. The results demonstrate its optimality, robustness, and sample efficiency over the other state-of-the-art decentralized MARL algorithms.
523 citations
TL;DR: In this paper, the authors identify the major challenges that autonomous cars are facing today is driving in urban environments and propose future research directions, including design approaches for autonomous vehicles that communicate with pedestrians and visual perception and reasoning algorithms tailored to understanding pedestrian intention.
Abstract: One of the major challenges that autonomous cars are facing today is driving in urban environments. To make it a reality, autonomous vehicles require the ability to communicate with other road users and understand their intentions. Such interactions are essential between vehicles and pedestrians, the most vulnerable road users. Understanding pedestrian behavior, however, is not intuitive and depends on various factors, such as demographics of the pedestrians, traffic dynamics, environmental conditions, and so on. In this paper, we identify these factors by surveying pedestrian behavior studies, both the classical works on pedestrian–driver interaction and the modern ones that involve autonomous vehicles. To this end, we will discuss various methods of studying pedestrian behavior and analyze how the factors identified in the literature are interrelated. We will also review the practical applications aimed at solving the interaction problem, including design approaches for autonomous vehicles that communicate with pedestrians and visual perception and reasoning algorithms tailored to understanding pedestrian intention. Based on our findings, we will discuss the open problems and propose future research directions.
391 citations
TL;DR: The main algorithms in motion planning, their features, and their applications to highway driving are reviewed, along with current and future challenges and open issues.
Abstract: Self-driving vehicles will soon be a reality, as main automotive companies have announced that they will sell their driving automation modes in the 2020s. This technology raises relevant controversies, especially with recent deadly accidents. Nevertheless, autonomous vehicles are still popular and attractive thanks to the improvement they represent to people’s way of life (safer and quicker transit, more accessible, comfortable, convenient, efficient, and environment-friendly). This paper presents a review of motion planning techniques over the last decade with a focus on highway planning. In the context of this article, motion planning denotes path generation and decision making. Highway situations limit the problem to high speed and small curvature roads, with specific driver rules, under a constrained environment framework. Lane change, obstacle avoidance, car following, and merging are the situations addressed in this paper. After a brief introduction to the context of autonomous ground vehicles, the detailed conditions for motion planning are described. The main algorithms in motion planning, their features, and their applications to highway driving are reviewed, along with current and future challenges and open issues.
333 citations
TL;DR: Yan et al. as mentioned in this paper proposed a novel network architecture, named feature pyramid and hierarchical boosting network (FPHBN), which integrates context information to low-level features for crack detection in a feature pyramid way, and it balances the contributions of both easy and hard samples to loss by nested sample reweighting in a hierarchical way during training.
Abstract: Pavement crack detection is a critical task for insuring road safety. Manual crack detection is extremely time-consuming. Therefore, an automatic road crack detection method is required to boost this progress. However, it remains a challenging task due to the intensity inhomogeneity of cracks and complexity of the background, e.g., the low contrast with surrounding pavements and possible shadows with a similar intensity. Inspired by recent advances of deep learning in computer vision, we propose a novel network architecture, named feature pyramid and hierarchical boosting network (FPHBN), for pavement crack detection. The proposed network integrates context information to low-level features for crack detection in a feature pyramid way, and it balances the contributions of both easy and hard samples to loss by nested sample reweighting in a hierarchical way during training. In addition, we propose a novel measurement for crack detection named average intersection over union (AIU). To demonstrate the superiority and generalizability of the proposed method, we evaluate it on five crack datasets and compare it with the state-of-the-art crack detection, edge detection, and semantic segmentation methods. The extensive experiments show that the proposed method outperforms these methods in terms of accuracy and generalizability. Code and data can be found in https://github.com/fyangneil/pavement-crack-detection .
307 citations
TL;DR: A comprehensive review of the state-of-the-art of deep learning-based approaches for vehicle behavior prediction based on three criteria: input representation, output type, and prediction method is provided.
Abstract: Behaviour prediction function of an autonomous vehicle predicts the future states of the nearby vehicles based on the current and past observations of the surrounding environment. This helps enhance their awareness of the imminent hazards. However, conventional behavior prediction solutions are applicable in simple driving scenarios that require short prediction horizons. Most recently, deep learning-based approaches have become popular due to their promising performance in more complex environments compared to the conventional approaches. Motivated by this increased popularity, we provide a comprehensive review of the state-of-the-art of deep learning-based approaches for vehicle behavior prediction in this article. We firstly give an overview of the generic problem of vehicle behavior prediction and discuss its challenges, followed by classification and review of the most recent deep learning-based solutions based on three criteria: input representation, output type, and prediction method. The article also discusses the performance of several well-known solutions, identifies the research gaps in the literature and outlines potential new research directions.
283 citations
TL;DR: A multi-stream feature fusion approach to extract and integrate rich features from traffic data and leverage a data-driven adjacent matrix instead of the distance-based matrix to construct graphs is proposed.
Abstract: Accurate and timely traffic flow prediction is crucial for intelligent transportation systems (ITS). Recent advances in graph-based neural networks have achieved promising prediction results. However, some challenges remain, especially regarding graph construction and the time complexity of models. In this paper, we propose a multi-stream feature fusion approach to extract and integrate rich features from traffic data and leverage a data-driven adjacent matrix instead of the distance-based matrix to construct graphs. We calculate the Spearman rank correlation coefficient between monitor stations to obtain the initial adjacent matrix and fine-tune it while training. As to the model, we construct a multi-stream feature fusion block (MFFB) module, which includes a three-channel network and the soft-attention mechanism. The three-channel networks are graph convolutional neural network (GCN), gated recurrent unit (GRU) and fully connected neural network (FNN), which are used to extract spatial, temporal and other features, respectively. The soft-attention mechanism is utilized to integrate the obtained features. The MFFB modules are stacked, and a fully connected layer and a convolutional layer are used to make predictions. We conduct experiments on two real-world traffic prediction tasks and verify that our proposed approach outperforms the state-of-the-art methods within an acceptable time complexity.
269 citations
TL;DR: An improved quantum evolutionary algorithm (QEA) based on the niche co-evolution strategy and enhanced particle swarm optimization (PSO) is designed, and an IPOQEA-based gate allocation method is proposed to allocate the flights to suitable gates within different periods.
Abstract: With the continuous and rapid growth of air traffic demand, gate resource becomes a major bottleneck restricting airport development. Rational gate allocation is regarded as one of the most important means to solve this bottleneck. In this paper, in order to comprehensively considere different stakeholders, a three-objective gate allocation model is to consider a wider scope, in which the minimizing passenger walking distances, the most balanced idle time of each gate and the best full use of large gate are optimized simultaneously to improve the practical efficiency. To efficiently solve this model, an improved quantum evolutionary algorithm (QEA) based on the niche co-evolution strategy and enhanced particle swarm optimization (PSO), namely IPOQEA is designed. An IPOQEA-based gate allocation method is proposed to allocate the flights to suitable gates within different periods. Finally, the actual operation data of Baiyun Airport is used to validate the effectiveness of the proposed method. Comparison results show that the constructed model can address the passenger walking distances, robustness and costs in airport management. Moreover, the IPOQEA has better optimization ability in solving gate allocation problem. Therefore, the proposed gate allocation method has great potential for practical engineering since it can easily make decisions for airport managers.
241 citations
TL;DR: A comprehensive review on the fault detection and diagnosis techniques for high-speed trains is presented using data-driven methods which are receiving increasing attention in transportation fields over the past ten years.
Abstract: High-speed trains have become one of the most important and advanced branches of intelligent transportation, of which the reliability and safety are still not mature enough for keeping up with other aspects. The first objective of this paper is to present a comprehensive review on the fault detection and diagnosis (FDD) techniques for high-speed trains. The second purpose of this work is, motivated by the pros and cons of the FDD methods for high-speed trains, to provide researchers and practitioners with informative guidance. Then, the application of FDD for high-speed trains is presented using data-driven methods which are receiving increasing attention in transportation fields over the past ten years. Finally, the challenges and promising issues are speculated for the future investigation.
239 citations
TL;DR: This survey extensively summarizes existing works in this field by categorizing them with respect to application types, control models and studied algorithms and discusses the challenges and open questions regarding deep RL-based transportation applications.
Abstract: Latest technological improvements increased the quality of transportation. New data-driven approaches bring out a new research direction for all control-based systems, e.g., in transportation, robotics, IoT and power systems. Combining data-driven applications with transportation systems plays a key role in recent transportation applications. In this paper, the latest deep reinforcement learning (RL) based traffic control applications are surveyed. Specifically, traffic signal control (TSC) applications based on (deep) RL, which have been studied extensively in the literature, are discussed in detail. Different problem formulations, RL parameters, and simulation environments for TSC are discussed comprehensively. In the literature, there are also several autonomous driving applications studied with deep RL models. Our survey extensively summarizes existing works in this field by categorizing them with respect to application types, control models and studied algorithms. In the end, we discuss the challenges and open questions regarding deep RL-based transportation applications.
TL;DR: This paper presents a survey that highlights the role modeling techniques within the realm of deep learning have played within ITS, focusing on how practitioners have formulated problems to address these various challenges, and outline both architectural and problem-specific considerations used to develop solutions.
Abstract: Transportation systems operate in a domain that is anything but simple. Many exhibit both spatial and temporal characteristics, at varying scales, under varying conditions brought on by external sources such as social events, holidays, and the weather. Yet, modeling the interplay of factors, devising generalized representations, and subsequently using them to solve a particular problem can be a challenging task. These situations represent only a fraction of the difficulties faced by modern intelligent transportation systems (ITS). In this paper, we present a survey that highlights the role modeling techniques within the realm of deep learning have played within ITS. We focus on how practitioners have formulated problems to address these various challenges, and outline both architectural and problem-specific considerations used to develop solutions. We hope this survey can help to serve as a bridge between the machine learning and transportation communities, shedding light on new domains and considerations in the future.
TL;DR: In this paper, the authors provide insight into the hierarchical motion planning problem and describe the basics of Deep Reinforcement Learning (DRL), and present state-of-the-art solutions systematized by different tasks and levels of autonomous driving, such as car-following, lane-keeping, trajectory following, merging, or driving in dense traffic.
Abstract: Academic research in the field of autonomous vehicles has reached high popularity in recent years related to several topics as sensor technologies, V2X communications, safety, security, decision making, control, and even legal and standardization rules. Besides classic control design approaches, Artificial Intelligence and Machine Learning methods are present in almost all of these fields. Another part of research focuses on different layers of Motion Planning, such as strategic decisions, trajectory planning, and control. A wide range of techniques in Machine Learning itself have been developed, and this article describes one of these fields, Deep Reinforcement Learning (DRL). The paper provides insight into the hierarchical motion planning problem and describes the basics of DRL. The main elements of designing such a system are the modeling of the environment, the modeling abstractions, the description of the state and the perception models, the appropriate rewarding, and the realization of the underlying neural network. The paper describes vehicle models, simulation possibilities and computational requirements. Strategic decisions on different layers and the observation models, e.g., continuous and discrete state representations, grid-based, and camera-based solutions are presented. The paper surveys the state-of-art solutions systematized by the different tasks and levels of autonomous driving, such as car-following, lane-keeping, trajectory following, merging, or driving in dense traffic. Finally, open questions and future challenges are discussed.
TL;DR: An integrated train operation approach by jointly optimizing the train timetable and driving strategy and a distributed regenerative braking energy model is proposed, based on which the integrated optimization model is formulated.
Abstract: Energy-efficient train operation is regarded as an effective way to reduce the operational cost and carbon emissions in metro systems. Reduction of the traction energy and increasing of the regenerative energy are two important ways for saving energy, which is closely related to the train timetable and driving strategy. To minimize the systematic net energy consumption, i.e., the difference between the traction energy consumption and the reused regenerative energy, this paper proposes an integrated train operation approach by jointly optimizing the train timetable and driving strategy. A precise train driving strategy is presented and the timetable model considers the headway between successive trains, the distribution of the trip time, and passenger demand in this paper. In addition, a distributed regenerative braking energy model is proposed, based on which the integrated optimization model is formulated. Then, a two-level approach is proposed to solve the problem. At the driving strategy level, the train control problem is transferred into a multi-step decision problem and the Dynamic Programming method is introduced to calculate the energy-efficient driving strategy with the given trip time. As for the timetable level, the trip times and headway of trains are optimized by using the Simulated Annealing algorithm based on the results of dynamic programming method. The timetable optimization level balances the mechanical traction energy of multi-interstations and the amount of the reused regenerative energy such that the net mechanical energy consumption of the metro system is minimized. Furthermore, two numerical examples are conducted for train operations in the peak and off-peak hours separately based on the real-world data of a metro line. The simulation results illustrate that the proposed approach can produce a good performance on energy-saving.
TL;DR: To ensure the reliability and safety of high-speed trains, detection and diagnosis of faults (FDD) in traction systems have become an active issue in the transportation area over the past two decades and most of the mainstream methods are reviewed.
Abstract: Recently, to ensure the reliability and safety of high-speed trains, detection and diagnosis of faults (FDD) in traction systems have become an active issue in the transportation area over the past two decades. Among these FDD methods, data-driven designs, that can be directly implemented without a logical or mathematical description of traction systems, have received special attention because of their overwhelming advantages. Based on the existing data-driven FDD methods for traction systems in high-speed trains, the first objective of this paper is to systematically review and categorize most of the mainstream methods. By analyzing the characteristic of observations from sensors equipped in traction systems, great challenges which may prevent successful FDD implementations on practical high-speed trains are then summarized in detail. Benefiting from theoretical developments of data-driven FDD strategies, instructive perspectives on this topic are further elaborately conceived by the integration of model-based FDD issues, system identification techniques, and new machine learning tools, which provide several promising solutions to FDD strategies for traction systems in high-speed trains.
TL;DR: The experimental and simulation results are presented to demonstrate the improved performance in tracking accuracy, steering smoothness, and computational efficiency compared to the MPC and the full-state feedback control.
Abstract: This paper presents a preview steering control algorithm and its closed-loop system analysis and experimental validation for accurate, smooth, and computationally inexpensive path tracking of automated vehicles. The path tracking issue is formulated as an optimal control problem with dynamic disturbance, i.e., the future road curvature. A discrete-time preview controller is then designed on the top of a linear augmented error system, in which the disturbances within a finite preview window are augmented as part of the state vector. The obtained optimal steering control law is in an analytic form and consists of two parts: 1) a feedback control responding to tracking errors and 2) a feedforward control dealing with the future road curvatures. The designed control’s nature, capacity, computation load, and underlying mechanism are revealed by the analysis of system responses in the time domain and the frequency domain, theoretical steady-state error, and comparison with the model predictive control (MPC). The algorithm was implemented on an automated vehicle platform, a hybrid Lincoln MKZ. The experimental and simulation results are then presented to demonstrate the improved performance in tracking accuracy, steering smoothness, and computational efficiency compared to the MPC and the full-state feedback control.
TL;DR: An IVN environment is introduced, and the constraints and characteristics of an intrusion detection system (IDS) design for IVNs are presented, and a survey of the proposed IDS designs for the IVNs is conducted.
Abstract: The development of the complexity and connectivity of modern automobiles has caused a massive rise in the security risks of in-vehicle networks (IVNs). Nevertheless, existing IVN designs (e.g., controller area network) lack cybersecurity consideration. Intrusion detection, an effective method for defending against cyberattacks on IVNs while providing functional safety and real-time communication guarantees, aims to address this issue. Therefore, the necessity of its research has risen. In this paper, an IVN environment is introduced, and the constraints and characteristics of an intrusion detection system (IDS) design for IVNs are presented. A survey of the proposed IDS designs for the IVNs is conducted, and the corresponding drawbacks are highlighted. Various optimization objectives are considered and comprehensively compared. Lastly, the trend, open issues, and emerging research directions are described.
TL;DR: A deep irregular convolutional residual LSTM network model called DST-ICRL is proposed which significantly outperforms both traditional and deep learning based urban traffic passenger flows prediction methods.
Abstract: Urban traffic passenger flows prediction is practically important to facilitate many real applications including transportation management and public safety. Recently, deep learning based approaches are proposed to learn the spatio-temporal characteristics of the traffic passenger flows. However, it is still very challenging to handle some complex factors such as hybrid transportation lines, mixed traffic, transfer stations, and some extreme weathers. Considering the multi-channel and irregularity properties of urban traffic passenger flows in different transportation lines, a more efficient and fine-grained deep spatio-temporal feature learning model is necessary. In this paper, we propose a deep irregular convolutional residual LSTM network model called DST-ICRL for urban traffic passenger flows prediction. We first model the passenger flows among different traffic lines in a transportation network into multi-channel matrices analogous to the RGB pixel matrices of an image. Then, we propose a deep learning framework that integrates irregular convolutional residential network and LSTM units to learn the spatial-temporal feature representations. To fully utilize the historical passenger flows, we sample both the short-term and long-term historical traffic data, which can capture the periodicity and trend of the traffic passenger flows. In addition, we also fuse other external factors further to facilitate a real-time prediction. We conduct extensive experiments on different types of traffic passenger flows datasets including subway, taxi and bus flows in Beijing as well as bike flows in New York. The results show that the proposed DST-ICRL significantly outperforms both traditional and deep learning based urban traffic passenger flows prediction methods.
TL;DR: A review of recent deep-learning-based data fusion approaches that leverage both image and point cloud data processing and identifies gaps and over-looked challenges between current academic researches and real-world applications.
Abstract: Autonomous vehicles were experiencing rapid development in the past few years. However, achieving full autonomy is not a trivial task, due to the nature of the complex and dynamic driving environment. Therefore, autonomous vehicles are equipped with a suite of different sensors to ensure robust, accurate environmental perception. In particular, the camera-LiDAR fusion is becoming an emerging research theme. However, so far there has been no critical review that focuses on deep-learning-based camera-LiDAR fusion methods. To bridge this gap and motivate future research, this article devotes to review recent deep-learning-based data fusion approaches that leverage both image and point cloud. This review gives a brief overview of deep learning on image and point cloud data processing. Followed by in-depth reviews of camera-LiDAR fusion methods in depth completion, object detection, semantic segmentation, tracking and online cross-sensor calibration, which are organized based on their respective fusion levels. Furthermore, we compare these methods on publicly available datasets. Finally, we identified gaps and over-looked challenges between current academic researches and real-world applications. Based on these observations, we provide our insights and point out promising research directions.
TL;DR: It is shown that CrackNet-V yields better overall performance particularly in detecting fine cracks compared with CrackNet, and further reveals the advantages of deep learning techniques for automated pixel-level pavement crack detection.
Abstract: A few recent developments have demonstrated that deep-learning-based solutions can outperform traditional algorithms for automated pavement crack detection. In this paper, an efficient deep network called CrackNet-V is proposed for automated pixel-level crack detection on 3D asphalt pavement images. Compared with the original CrackNet, CrackNet-V has a deeper architecture but fewer parameters, resulting in improved accuracy and computation efficiency. Inspired by CrackNet, CrackNet-V uses invariant spatial size through all layers such that supervised learning can be conducted at pixel level. Following the VGG network, CrackNet-V uses $3\times 3$ size of filters for the first six convolutional layers and stacks several $3\times 3$ convolutional layers together for deep abstraction, resulting in reduced number of parameters and efficient feature extraction. CrackNet-V has 64113 parameters and consists of ten layers, including one pre-process layer, eight convolutional layers, and one output layer. A new activation function leaky rectified tanh is proposed in this paper for higher accuracy in detecting shallow cracks. The training of CrackNet-V was completed after 3000 iterations, which took only one day on a GeForce GTX 1080Ti device. According to the experimental results on 500 testing images, CrackNet-V achieves a high performance with a Precision of 84.31%, Recall of 90.12%, and an F-1 score of 87.12%. It is shown that CrackNet-V yields better overall performance particularly in detecting fine cracks compared with CrackNet. The efficiency of CrackNet-V further reveals the advantages of deep learning techniques for automated pixel-level pavement crack detection.
TL;DR: A convolutional neural network approach, the mask R-CNN, is adopted to address the full pipeline of detection and recognition with automatic end-to-end learning, which is sufficient for deployment in practical applications of the traffic-sign inventory management.
Abstract: Automatic detection and recognition of traffic signs plays a crucial role in management of the traffic-sign inventory. It provides an accurate and timely way to manage traffic-sign inventory with a minimal human effort. In the computer vision community, the recognition and detection of traffic signs are a well-researched problem. A vast majority of existing approaches perform well on traffic signs needed for advanced driver-assistance and autonomous systems. However, this represents a relatively small number of all traffic signs (around 50 categories out of several hundred) and performance on the remaining set of traffic signs, which are required to eliminate the manual labor in traffic-sign inventory management, remains an open question. In this paper, we address the issue of detecting and recognizing a large number of traffic-sign categories suitable for automating traffic-sign inventory management. We adopt a convolutional neural network (CNN) approach, the mask R-CNN, to address the full pipeline of detection and recognition with automatic end-to-end learning. We propose several improvements that are evaluated on the detection of traffic signs and result in an improved overall performance. This approach is applied to detection of 200 traffic-sign categories represented in our novel dataset. The results are reported on highly challenging traffic-sign categories that have not yet been considered in previous works. We provide comprehensive analysis of the deep learning method for the detection of traffic signs with a large intra-category appearance variation and show below 3% error rates with the proposed approach, which is sufficient for deployment in practical applications of the traffic-sign inventory management.
TL;DR: This paper presents visions and works on integrating the artificial intelligent transportation systems and the real intelligent Transportation systems to create and enhance “intelligence” of IoT-enabled ITS, and presents some case studies to demonstrate the effectiveness of parallel transportation systems.
Abstract: IoT-driven intelligent transportation systems (ITS) have great potential and capacity to make transportation systems efficient, safe, smart, reliable, and sustainable. The IoT provides the access and driving forces of seamlessly integrating transportation systems from the physical world to the virtual counterparts in the cyber world. In this paper, we present visions and works on integrating the artificial intelligent transportation systems and the real intelligent transportation systems to create and enhance “intelligence” of IoT-enabled ITS. With the increasing ubiquitous and deep sensing capacity of IoT-enabled ITS, we can quickly create artificial transportation systems equivalent to physical transportation systems in computers, and thus have parallel intelligent transportation systems, i.e. the real intelligent transportation systems and artificial intelligent transportation systems. The evolution process of transportation system is studied in the view of the parallel world. We can use a large number of long-term iterative simulation to predict and analyze the expected results of operations. Thus, truly effective and smart ITS can be planned, designed, built, operated and used. The foundation of the parallel intelligent transportation systems is based on the ACP theory, which is composed of artificial societies, computational experiments, and parallel execution. We also present some case studies to demonstrate the effectiveness of parallel transportation systems.
TL;DR: The result shows that by specifying an effective reward function, a controller can be learned and works well under different traffic demands as well as traffic light cycles with different durations.
Abstract: The concept of Connected and Automated Vehicles (CAVs) enables instant traffic information to be shared among vehicle networks. With this newly proposed concept, a vehicle’s driving behaviour will no longer be solely based on the driver’s limited and incomplete observation. By taking advantages of the shared information, driving behaviours of CAVs can be improved greatly to a more responsible, accurate and efficient level. This study proposed a reinforcement-learning-based car following model for CAVs in order to obtain an appropriate driving behaviour to improve travel efficiency, fuel consumption and safety at signalized intersections in real-time. The result shows that by specifying an effective reward function, a controller can be learned and works well under different traffic demands as well as traffic light cycles with different durations. This study reveals a great potential of emerging reinforcement learning technologies in transport research and applications.
TL;DR: The proposed end-to-end Line-CNN system, in which the key component is a novel line proposal unit (LPU), outperforms the state-of-the-art methods, indicates the practicability and effectiveness of L-CNN for real-time intelligent self-driving systems.
Abstract: The task of traffic line detection is a fundamental yet challenging problem. Previous approaches usually conduct traffic line detection via a two-stage way, namely the line segment detection followed by a segment clustering, which is very likely to ignore the global semantic information of an entire line. To address the problem, we propose an end-to-end system called Line-CNN (L-CNN), in which the key component is a novel line proposal unit (LPU). The LPU utilizes line proposals as references to locate accurate traffic curves, which forces the system to learn the global feature representation of the entire traffic lines. We benchmark the proposed L-CNN on two public datasets including MIKKI and TuSimple, and the results suggest that L-CNN outperforms the state-of-the-art methods. In addition, L-CNN can run at approximately 30 f/s on a Titan X GPU, which indicates the practicability and effectiveness of L-CNN for real-time intelligent self-driving systems.
TL;DR: A method for predicting driver intent at urban intersections through multi-modal trajectory prediction with uncertainty based on recurrent neural networks combined with a mixture density network output layer and a clustering algorithm.
Abstract: Understanding the intentions of drivers at intersections is a critical component for autonomous vehicles. Urban intersections that do not have traffic signals are a common epicenter of highly variable vehicle movement and interactions. We present a method for predicting driver intent at urban intersections through multi-modal trajectory prediction with uncertainty. Our method is based on recurrent neural networks combined with a mixture density network output layer. To consolidate the multi-modal nature of the output probability distribution, we introduce a clustering algorithm that extracts the set of possible paths that exist in the prediction output and ranks them according to probability. To verify the method’s performance and generalizability, we present a real-world dataset that consists of over 23 000 vehicles traversing five different intersections, collected using a vehicle-mounted lidar-based tracking system. An array of metrics is used to demonstrate the performance of the model against several baselines.
TL;DR: This paper combines Monte Carlo tree search (MCTS) and some heuristic rules to find a nearly global-optimal passing order (leaf node) within a very short planning time.
Abstract: In this paper, we propose a new cooperative driving strategy for connected and automated vehicles (CAVs) at unsignalized intersections. Based on the tree representation of the solution space for the passing order, we combine Monte Carlo tree search (MCTS) and some heuristic rules to find a nearly global-optimal passing order (leaf node) within a very short planning time. Testing results show that this new strategy can keep a good tradeoff between performance and computation flexibility.
TL;DR: This paper formulates the EVCS problem as a hierarchical mixed-variable optimization problem, considering the dependency among the station selection, the charging option at each station and the charging amount settings, and specifically design a Mixed-Variable Differentiate Evolution (MVDE) as the scheduling algorithm for this problem.
Abstract: The increasing popularity of battery-limited electric vehicles puts forward an important issue of how to charge the vehicles effectively. This problem, commonly referred to as Electric Vehicle Charging Scheduling (EVCS), has been proven to be NP-hard. Most of the existing works formulate the EVCS problem simply as a constrained shortest path finding problem and treat it by discrete optimization. However, other variables such as the charging amount of energy and the charging option at a station need to be considered in practical use. This paper hence formulates the EVCS problem as a hierarchical mixed-variable optimization problem, considering the dependency among the station selection, the charging option at each station and the charging amount settings. To adapt to the new problem model, we specifically design a Mixed-Variable Differentiate Evolution (MVDE) as the scheduling algorithm for our proposed EVCS system. The MVDE contains several specific operators, including a charging station route construction, a hierarchical mixed-variable mutation operator and a constraint-aware evaluation operator. Experimental results validate the effectiveness of our proposed MVDE-based system on both synthetic and real-world transportation networks.
TL;DR: This paper proposes a context-aware reliable beaconing scheme, called CoBe, to enhance the broadcast reliability for safety applications, and conducts intensive data analytics on V2V performance, based on a large amount of real-world DSRC communications trace collected in Shanghai city.
Abstract: The IEEE 802.11p-based dedicated short range communication (DSRC) is essential to enhance driving safety and improve road efficiency by enabling rapid cooperative message exchanging. However, there is a lack of good understanding on the DSRC performance in urban environments for vehicle-to-vehicle (V2V) communications, which impedes its reliable and efficient application. In this paper, we first conduct intensive data analytics on V2V performance, based on a large amount of real-world DSRC communications trace collected in Shanghai city, and obtain several key insights as follows. First, among many context factors, the non-line-of-sight (NLoS) link condition is the major factor degrading V2V performance. Second, the durations of line-of-sight (LoS) and NLoS transmission conditions follow power law distributions, which indicate that the probability of experiencing long LoS/NLoS conditions both could be high. Third, the packet inter-reception (PIR) time distribution follows an exponential distribution in the LoS conditions but a power law in the NLoS conditions, which means that the consecutive packet reception failures rarely appear in the LoS conditions but can constantly appear in the NLoS conditions. Based on these findings, we propose a context-aware reliable beaconing scheme, called CoBe , to enhance the broadcast reliability for safety applications. The CoBe is a fully distributed scheme, in which a vehicle first detects the link condition with each of its neighbors by machine learning algorithms, then exchanges such link condition information with its neighbors, and finally selects the minimal number of helper vehicles to rebroadcast its beacons to those neighbors in bad link condition. To analyze and evaluate the CoBe performance, a two-state Markov chain is devised to model beaconing behaviors. The extensive trace-driven simulations are conducted to demonstrate the efficacy of CoBe .
TL;DR: In this article, an end-to-end autonomous driving system using RGB and depth modalities is proposed, which uses early, mid and late fusion schemes, both in multisensory and single-sensor (monocular depth estimation) settings.
Abstract: A crucial component of an autonomous vehicle (AV) is the artificial intelligence (AI) is able to drive towards a desired destination. Today, there are different paradigms addressing the development of AI drivers. On the one hand, we find modular pipelines, which divide the driving task into sub-tasks such as perception and maneuver planning and control. On the other hand, we find end-to-end driving approaches that try to learn a direct mapping from input raw sensor data to vehicle control signals. The later are relatively less studied, but are gaining popularity since they are less demanding in terms of sensor data annotation. This paper focuses on end-to-end autonomous driving. So far, most proposals relying on this paradigm assume RGB images as input sensor data. However, AVs will not be equipped only with cameras, but also with active sensors providing accurate depth information (e.g., LiDARs). Accordingly, this paper analyses whether combining RGB and depth modalities, i.e. using RGBD data, produces better end-to-end AI drivers than relying on a single modality. We consider multimodality based on early, mid and late fusion schemes, both in multisensory and single-sensor (monocular depth estimation) settings. Using the CARLA simulator and conditional imitation learning (CIL), we show how, indeed, early fusion multimodality outperforms single-modality.
TL;DR: Wang et al. as discussed by the authors proposed quadruple directional deep learning (QD-DLF) for improving vehicle re-identification performance, which is based on the same basic deep learning architecture that is a shortly and densely connected convolutional neural network.
Abstract: In order to resist the adverse effect of viewpoint variations, we design quadruple directional deep learning networks to extract quadruple directional deep learning features (QD-DLF) of vehicle images for improving vehicle re-identification performance. The quadruple directional deep learning networks are of similar overall architecture, including the same basic deep learning architecture but different directional feature pooling layers. Specifically, the same basic deep learning architecture that is a shortly and densely connected convolutional neural network is utilized to extract the basic feature maps of an input square vehicle image in the first stage. Then, the quadruple directional deep learning networks utilize different directional pooling layers, i.e., horizontal average pooling layer, vertical average pooling layer, diagonal average pooling layer, and anti-diagonal average pooling layer, to compress the basic feature maps into horizontal, vertical, diagonal, and anti-diagonal directional feature maps, respectively. Finally, these directional feature maps are spatially normalized and concatenated together as a quadruple directional deep learning feature for vehicle re-identification. The extensive experiments on both VeRi and VehicleID databases show that the proposed QD-DLF approach outperforms multiple state-of-the-art vehicle re-identification methods.