相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Computer Science, Theory & Methods
Chao Yu et al.
Summary: As a subfield of machine learning, reinforcement learning aims to optimize decision making by using interaction samples of an agent with its environment and the potentially delayed feedbacks. It tackles sequential decision-making problems with sampled, evaluative, and delayed feedbacks simultaneously. In healthcare domains, it is a suitable candidate for developing powerful solutions in various areas such as treatment regimes, medical diagnosis, and control or scheduling problems. Challenges, open issues, potential solutions, and future research directions are also discussed.
ACM COMPUTING SURVEYS
(2023)
Article
Computer Science, Artificial Intelligence
Ngan Le et al.
Summary: This work provides a comprehensive review of recent and state-of-the-art research advances of deep reinforcement learning in computer vision. It categorizes deep reinforcement learning methodologies and discusses their advantages and limitations in various applications in computer vision. The analysis includes reinforcement learning techniques, network design, performance evaluation, publicly available datasets, source code availability, and future research directions.
ARTIFICIAL INTELLIGENCE REVIEW
(2022)
Article
Engineering, Multidisciplinary
Zina Chkirbene et al.
Summary: Smart health systems can enhance our quality of life by integrating diverse technologies and information. This paper proposes a novel Deep Reinforcement Learning (DRL) model that addresses the challenging network selection problem in dense heterogeneous networks (HetNets) over 5G network. The model optimizes medical data delivery by minimizing transmission energy consumption and latency while meeting the Quality of Service (QoS) requirements. Experimental results demonstrate that the proposed DRL-based model outperforms greedy techniques, achieving minimized energy consumption and latency while satisfying different users' demands in dynamic environments.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2022)
Article
Engineering, Civil
Jianyu Chen et al.
Summary: This article introduces an interpretable deep reinforcement learning method for end-to-end autonomous driving, which utilizes a sequential latent environment model for handling complex urban scenarios and significantly reducing the sample complexity of reinforcement learning. Comparative tests in a realistic driving simulator demonstrate that the method outperforms many baseline models including DQN, DDPG, TD3, and SAC in crowded urban environments.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
B. Ravi Kiran et al.
Summary: This paper summarizes deep reinforcement learning algorithms, provides a taxonomy of automated driving tasks, discusses key computational challenges in real world deployment of autonomous driving agents, and explores adjacent domains as well as the role of simulators in training agents.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
Marwa Ahmed et al.
Summary: The paper proposes a hybrid framework with affordance learning to simplify urban scenes for training an AV agent, along with a planned trajectory and driving measurements. It involves a supervised learning network for affordance predictions and a deep deterministic policy gradient model for mapping perception information into actions. The proposed solution generalizes well in different traffic solutions and outperforms other baseline methods in handling various AV driving tasks with differing levels of difficulty.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Proceedings Paper
Computer Science, Information Systems
Yasser H. Khalil et al.
Summary: This paper proposes LidNet to enhance the accuracy of perception and motion prediction for scene understanding in autonomous driving. LidNet redesigns the MotionNet architecture by replacing spatial convolution layers with residual blocks and using average pooling instead of strided convolution for spatial reduction. The connections between encoders and decoders are adjusted to create a symmetric network. LidNet outperforms the state-of-the-art and operates in real-time according to extensive evaluation on the nuScenes dataset.
2022 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Sudeep Fadadu et al.
Summary: We propose an end-to-end method for object detection and trajectory prediction using multi-view representations of LiDAR returns and camera images. By recognizing the strengths and weaknesses of different view representations, we introduce an efficient and generic fusion method that combines the benefits of all views. Our approach extracts and fuses features from LiDAR data, high-definition maps, and raw LiDAR information to output final detections and trajectories. The proposed multi-view fusion approach achieves improved performance on real-world and public datasets with minimal increase in computational cost.
2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022)
(2022)
Article
Computer Science, Artificial Intelligence
Yasser H. Khalil et al.
Summary: Researchers have recently devoted considerable effort towards enhancing perception and motion prediction models, but often ignore using different sensor modalities for pixel-wise joint perception and motion prediction. Utilizing a camera sensor results in a substantial gain in perception and motion prediction, especially for small and distant objects.
IEEE OPEN JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Automation & Control Systems
Zhixiao Sun et al.
Summary: The researchers proposed a novel Multi-Agent Hierarchical Policy Gradient algorithm (MAHPG), capable of learning various strategies and surpassing expert cognition through adversarial self-play learning. The algorithm adopts a hierarchical decision network to handle complex and hybrid actions, similar to human decision-making ability, effectively reducing action ambiguity. Experimental results demonstrate that MAHPG excels in defense and offense ability compared to state-of-the-art air combat methods.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Robotics
Xieyuanli Chen et al.
Summary: In this letter, a novel approach for moving object segmentation from 3D LiDAR scans was proposed and compared to other state-of-the-art methods, showing superior segmentation quality. A new benchmark for LiDAR-based moving object segmentation was created based on SemanticKITTI, with the code published for transparency and further research comparison.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Engineering, Electrical & Electronic
Hongbo Zou et al.
Summary: This article investigates the optimal management of a multi-carrier water and energy system with high penetration of renewable energy sources, proposing a stochastic approach based on unscented transform to handle forecast error in energy demands and market prices. A novel reinforcement learning-based approach is devised to find near optimal solutions and reduce computational burden. Simulation results on a practical smart island show the effectiveness of the proposed model and its high efficacy.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Tanmay Agarwal et al.
Summary: The study combines modular and Deep Reinforcement Learning (DRL) methods to address planning and control problems in urban driving, designs a specific input representation to help the DRL agent learn urban driving tasks, and achieves state-of-the-art performance.
2021 IEEE INTELLIGENT TRANSPORTATION SYSTEMS CONFERENCE (ITSC)
(2021)
Proceedings Paper
Engineering, Electrical & Electronic
Yasser H. Khalil et al.
Summary: In this study, an end-to-end multi-view fusion methodology applied to MotionNet was proposed to enhance the accuracy of perception and motion prediction in autonomous vehicles. Experimental results demonstrated that our proposed extension using the multi-view fusion technique outperforms MotionNet, especially for small and distant objects in perception and motion prediction.
2021 IEEE 94TH VEHICULAR TECHNOLOGY CONFERENCE (VTC2021-FALL)
(2021)
Article
Computer Science, Information Systems
Yasser H. Khalil et al.
Summary: The paper introduces LiCaNext, a system that employs multi-modal fusion to improve accuracy in perception and motion prediction for autonomous driving. By incorporating sequential range residual images, it enhances performance with minimal impact on inference time, demonstrating significant gains in motion prediction.
Article
Computer Science, Artificial Intelligence
Abdoulaye O. Ly et al.
Summary: There is a growing interest in autonomous vehicles due to recent advances in machine learning and deep learning. The gap towards full autonomy is gradually diminishing with approaches like modular systems, direct perception techniques, and end-to-end frameworks. This paper aims to review recent works on end-to-end deep learning models for lane stable driving and available datasets for evaluation.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2021)
Proceedings Paper
Computer Science, Hardware & Architecture
Yasser H. Khalil et al.
Summary: The paper discusses the development and research focus of self-driving vehicles, proposing an autonomous navigation method that combines motion prediction and reinforcement learning to enhance traffic efficiency and safety. Experimental results show that performance can be further improved based on motion prediction.
IWCMC 2021: 2021 17TH INTERNATIONAL WIRELESS COMMUNICATIONS & MOBILE COMPUTING CONFERENCE (IWCMC)
(2021)
Article
Robotics
Sorin Grigorescu et al.
JOURNAL OF FIELD ROBOTICS
(2020)
Proceedings Paper
Automation & Control Systems
Simon Hecker et al.
2020 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2020)
Article
Engineering, Multidisciplinary
Ehtesham Hassan et al.
JOURNAL OF ENGINEERING
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Felipe Codevilla et al.
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019)
(2019)
Proceedings Paper
Automation & Control Systems
Alex Kendall et al.
2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA)
(2019)
Article
Computer Science, Artificial Intelligence
Yuanzhe Wang et al.
JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS
(2017)
Article
Multidisciplinary Sciences
Volodymyr Mnih et al.
