IEEE Transactions on Intelligent Vehicles最新文献_第7页

Panoptic Driving Perception Model and Inference Acceleration Based on FPGA 基于FPGA的全景驾驶感知模型及推理加速

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462431

Yang Yang;Hui Xu;Abdullah Aman Khan;Jie Shao

{"title":"Panoptic Driving Perception Model and Inference Acceleration Based on FPGA","authors":"Yang Yang;Hui Xu;Abdullah Aman Khan;Jie Shao","doi":"10.1109/TIV.2024.3462431","DOIUrl":"https://doi.org/10.1109/TIV.2024.3462431","url":null,"abstract":"Panoptic perception systems are critical for autonomous driving, as they process multiple visual tasks simultaneously, enhancing functionality. Compared with designing multiple independent networks to address various tasks, these systems exhibit reduced overall inference latency by integrating various tasks into a single network. Existing panoptic perception networks often rely on pre-trained classification models as their backbone, which are not tailored for specific tasks, thereby compromising accuracy. To address this, we propose a dual-branch backbone and a wide perception segmentation head, enhancing the effectiveness of the network for autonomous driving applications. This enhanced network can simultaneously perform vehicle object detection, drivable area segmentation, and lane segmentation. Furthermore, to meet the stringent latency requirements of autonomous driving, we implement this network using an FPGA acceleration card. In our experiments using the challenging BDD100K dataset, the model significantly surpasses the baseline in accuracy for all tasks. To satisfy the increased real-time demands, the VCK5000 FPGA is used, which achieves inference speeds approximately 35.2 times faster than GPU-based deployments and about 41.5 times energy efficiency, providing significant advantages in resource-constrained scenarios.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 6","pages":"3697-3704"},"PeriodicalIF":14.3,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Scalable Input-to-State Stability of Heterogeneous Systems With Applications in Vehicle Platooning 异构系统的可伸缩输入状态稳定性及其在车辆队列中的应用

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462511

Meng Qiu;Di Liu;Simone Baldi;Guodong Yin;Wenwu Yu;Ming Cao

引用次数: 0

Circular Formation Control of Target Enclosing for Fixed-Wing UAVs in Three-Dimensional Space 三维空间固定翼无人机目标封闭的圆形编队控制

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462457

Xiuhui Peng;Renkai Yi;Peng Wang;Yuezu Lv

引用次数: 0

Rethinking Early-Fusion Strategies for Improved Multispectral Object Detection 改进多光谱目标检测的早期融合策略再思考

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462488

Xue Zhang;Si-Yuan Cao;Fang Wang;Runmin Zhang;Zhe Wu;Xiaohan Zhang;Xiaokai Bai;Hui-Liang Shen

{"title":"Rethinking Early-Fusion Strategies for Improved Multispectral Object Detection","authors":"Xue Zhang;Si-Yuan Cao;Fang Wang;Runmin Zhang;Zhe Wu;Xiaohan Zhang;Xiaokai Bai;Hui-Liang Shen","doi":"10.1109/TIV.2024.3462488","DOIUrl":"https://doi.org/10.1109/TIV.2024.3462488","url":null,"abstract":"Most recent multispectral object detectors employ a two-branch structure to extract features from RGB and thermal images. While the two-branch structure achieves better performance than a single-branch structure, it overlooks inference efficiency. This conflict is increasingly aggressive, as recent works solely pursue higher performance rather than both performance and efficiency. In this paper, we address this issue by improving the performance of efficient single-branch structures. We revisit the reasons causing the performance gap between these structures. For the first time, we reveal the information interference problem in the naive early-fusion strategy adopted by previous single-branch structures. Besides, we find that the domain gap between multispectral images, and weak feature representation of the single-branch structure are also key obstacles for performance. Focusing on these three problems, we propose corresponding solutions, including a novel shape-priority early-fusion strategy, a weakly supervised learning method, and a core knowledge distillation technique. Experiments demonstrate that single-branch networks equipped with these three contributions achieve significant performance enhancements while retaining high efficiency.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 6","pages":"3728-3742"},"PeriodicalIF":14.3,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Observer-Based Pole-Zero Cancellation Trajectory-Tracking Control for Two-Wheeled Vehicles With Performance Recovery Property 具有性能恢复特性的两轮车辆基于观测器的极零抵消轨迹跟踪控制

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462445

Yonghun Kim;Sun Lim;Hyun Ho Kang;Seok-Kyoon Kim;Choon Ki Ahn;Ramesh K. Agarwal

引用次数: 0

Human-Inspired Autonomous Racing in Low Friction Environments 低摩擦环境中人类启发的自动驾驶赛车

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-17 DOI: 10.1109/TIV.2024.3462253

Trey P. Weber;Rajan K. Aggarwal;J. Christian Gerdes

{"title":"Human-Inspired Autonomous Racing in Low Friction Environments","authors":"Trey P. Weber;Rajan K. Aggarwal;J. Christian Gerdes","doi":"10.1109/TIV.2024.3462253","DOIUrl":"https://doi.org/10.1109/TIV.2024.3462253","url":null,"abstract":"Low friction surfaces such as ice and snow pose a significant challenge for autonomous vehicle maneuvering due to the limited available traction. While some prior works have successfully controlled a vehicle up to the friction limit, generalized autonomous driving in these conditions remains an open problem. Skilled human drivers, however, display exceptional vehicle control in these challenging environments. In particular, rally drivers operate with large sideslip angles and high tire slip to achieve their objective of minimizing time. We directly compare a professional human driver and state-of-the-art autonomous racing controller to reveal dramatic differences in state- and input- space utilization. Inspired by these ideas, we present a motion planning and control framework for autonomous racing in low friction environments. We first develop a vehicle model to capture the dynamics associated with high slip maneuvering. Then, we implement nonlinear model predictive control for online, time optimal trajectory generation. Experimental validation on a frozen lake testing track using a Volkswagen Golf GTI reveals the ability to safely operate beyond limits of conventional vehicle safety systems. The utility of this increased maneuverability is demonstrated when the controller is able to recover from a significant, unexpected disturbance to the yaw dynamics. These results suggest that autonomous vehicles, like skilled human drivers, can leverage increased slip to improve robustness in low friction environments.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 6","pages":"3684-3696"},"PeriodicalIF":14.3,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10681660","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

CoPreMo: A Collaborative Predictive Model in Time Series and Its Application to Radar Target Tracking for ADAS/AD Vehicles 时间序列协同预测模型CoPreMo及其在ADAS/AD车辆雷达目标跟踪中的应用

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-16 DOI: 10.1109/TIV.2024.3461730

Zie Eya Ekolle;Ryuji Kohno

{"title":"CoPreMo: A Collaborative Predictive Model in Time Series and Its Application to Radar Target Tracking for ADAS/AD Vehicles","authors":"Zie Eya Ekolle;Ryuji Kohno","doi":"10.1109/TIV.2024.3461730","DOIUrl":"https://doi.org/10.1109/TIV.2024.3461730","url":null,"abstract":"Automotivedriving assistant systems (ADAS) and Automated driving (AD) technologies are commonly employed in the control of unmanned vehicles along a path. However, like many other technologies, they come with risks, including potential misdirection and collisions with obstacles along the vehicle's route. To mitigate these risks, various tracking systems, including radar tracking systems, are employed to detect and monitor targets along the trajectories of ADAS/AD vehicles. Nevertheless, the effectiveness of these tracking operations is crucial in assessing the reliability of both tracking systems and ADAS/AD technologies, especially at the edge-computing level. In this study, we introduce a tracking technique using a collaborative predictive model in a time series, named CoPreMo, aimed at enhancing the reliability of radar system tracking operations. We conducted three experiments with this model on a simulated radar system to track a target's range at varying speeds across three ADAS/AD scenarios. The experiments yielded range tracking errors of 0.21 m, 0.26 m, and 0.32 m, outperforming the baseline models.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 6","pages":"3659-3669"},"PeriodicalIF":14.3,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255941","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ALGPT: Multi-Agent Cooperative Framework for Open-Vocabulary Multi-Modal Auto-Annotating in Autonomous Driving 自动驾驶中开放词汇多模态自动标注的多智能体协作框架

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-16 DOI: 10.1109/TIV.2024.3461651

Yijie Zhou;Xianhui Cheng;Qiming Zhang;Lei Wang;Wenchao Ding;Xiangyang Xue;Chunbo Luo;Jian Pu

{"title":"ALGPT: Multi-Agent Cooperative Framework for Open-Vocabulary Multi-Modal Auto-Annotating in Autonomous Driving","authors":"Yijie Zhou;Xianhui Cheng;Qiming Zhang;Lei Wang;Wenchao Ding;Xiangyang Xue;Chunbo Luo;Jian Pu","doi":"10.1109/TIV.2024.3461651","DOIUrl":"https://doi.org/10.1109/TIV.2024.3461651","url":null,"abstract":"Large Language Models (LLMs) have achieved impressive progress in decision-making and task automation for intelligent agents. However, multiple agents must cooperate to complete tasks in complex real-world applications, such as auto-annotating in autonomous driving. The primary challenges lie in how multiple agents effectively communicate and collaborate in a multi-modal environment and how to automatically refine annotating results to reduce human intervention. These challenges also hinder LLMs from fully evolving into embodied intelligent agents. Driven by these motivations, we propose ALGPT, a multi-agent cooperative framework for open-vocabulary multi-modal auto-annotation in autonomous driving. ALGPT dynamically assembles agent teams with different roles, and agents cooperate to complete annotation tasks according to requirements. By leveraging Chain of Thought (CoT) and In-Context Learning (ICL) techniques, ALGPT's reasoning capabilities are enhanced, allowing it to develop suitable plans autonomously without human intervention. Furthermore, drawing from project management standards, we introduce project management documents and Standard Operating Procedures (SOPs), which further align ALGPT's behavior with human expectations and mitigate the impact of GPT illusions caused by the cascading effects of multiple GPTs.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 5","pages":"3644-3658"},"PeriodicalIF":14.3,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards Few-Shot Object Detection Through Dual Calibration 基于双校准的少镜头目标检测

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-16 DOI: 10.1109/TIV.2024.3461742

Ding Sheng Ong;Yi Liu;Jungong Han

{"title":"Towards Few-Shot Object Detection Through Dual Calibration","authors":"Ding Sheng Ong;Yi Liu;Jungong Han","doi":"10.1109/TIV.2024.3461742","DOIUrl":"https://doi.org/10.1109/TIV.2024.3461742","url":null,"abstract":"Object detection is crucial in traffic scenes for accurately identifying multiple objects within complex environments. Traditional systems rely on deep learning models trained on large-scale datasets, but this approach can be expensive and impractical. Few-shot object detection (FSOD) offers a potential solution by addressing limited data availability. However, object detectors trained with FSOD frameworks often generalize poorly on classes with limited samples. Although most existing methods alleviate this problem by calibrating either the feature maps or prediction heads of the object detector, none of them, like this work, have proposed a unified, dual calibration strategy that operates in both the latent feature space and the prediction probability space of the object detector. Specifically, we propose to improve representation precision by reducing the variances of feature vectors using highly adaptive centroids learned from ensembles of training features in the latent space. These centroids are employed to calibrate the features and reveal the underlying structure of the latent feature space. Moreover, we further exploit the association between the query and support features to calibrate inaccurate predictions resulting from overfitting or underfitting when fine-tuned with few training samples and low training iterations. Through visualization, we demonstrate that our method produces more discriminative high-level features, ultimately improving the precision of an object detector's predictions. To validate the effectiveness of our approaches, we conduct comprehensive experiments on well-known benchmarks, including PASCAL VOC and MS-COCO, showing considerable performance gains compared to existing works.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 6","pages":"3670-3683"},"PeriodicalIF":14.3,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145255972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Reinforcement Learning Control for a Class of Discrete-Time Non-Strict Feedback Multi-Agent Systems and Application to Multi-Marine Vehicles 一类离散时间非严格反馈多智能体系统的强化学习控制及其在多船舶上的应用

IF 14.3 1区工程技术

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-09-11 DOI: 10.1109/TIV.2024.3458894

Weiwei Bai;Dewang Chen;Bo Zhao;Andrea D'Ariano

{"title":"Reinforcement Learning Control for a Class of Discrete-Time Non-Strict Feedback Multi-Agent Systems and Application to Multi-Marine Vehicles","authors":"Weiwei Bai;Dewang Chen;Bo Zhao;Andrea D'Ariano","doi":"10.1109/TIV.2024.3458894","DOIUrl":"https://doi.org/10.1109/TIV.2024.3458894","url":null,"abstract":"A novel control design problem for a class of non-strict feedback multi-agent systems (MAS) in discrete-time form is studied based on reinforcement learning (RL) and applied to multi-marine vehicles (MMV). Firstly, for this kind of discrete-time MAS, a novel system transformation, which can not only solve the noncausal problem that exists in the backstepping method but also reduce the computational complexity, is proposed. Secondly, the algebraic-loop problem inherent in the conventional controller design is solved by compensating the dynamics and using the property of neural network (NN). Thirdly, the multi-gradient recursive (MGR) RL scheme is developed for the sake of designing the optimal controller. Finally, the stability analysis is presented, and all signals are ensured to be semi-global uniformly ultimately bounded (SGUUB) in the Lyapunov's sense. Besides, this scheme is applied to the MMV which can be described in the non-strict feedback form to extend the application of the designed controller. The MMV simulation demonstrates the validation of this scheme.","PeriodicalId":36532,"journal":{"name":"IEEE Transactions on Intelligent Vehicles","volume":"10 5","pages":"3613-3625"},"PeriodicalIF":14.3,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0