79,960 Hits in 4.0 sec

Action-Based Representation Learning for Autonomous Driving [article]

Yi Xiao, Felipe Codevilla, Christopher Pal, Antonio M. Lopez
2020 arXiv   pre-print
Alternatively, we propose to use this kind of action-based driving data for learning representations.  ...  Unfortunately, seemingly straightforward approaches for creating end-to-end driving models that map sensor data directly into driving actions are problematic in terms of interpretability, and typically  ...  López acknowledge the financial support received for this work from the Spanish TIN2017-88709-R (MINECO/AEI/FEDER, UE) project.  ... 
arXiv:2008.09417v2 fatcat:tqcvaaoswjbfrie63emjdidnpq

Vision-Based Autonomous Driving: A Model Learning Approach [article]

Ali Baheri, Ilya Kolmanovsky, Anouck Girard, H. Eric Tseng, and Dimitar Filev
2020 arXiv   pre-print
Finally, we utilize an evolutionary-based reinforcement learning algorithm to train a controller based on these latent representations to identify the action to take.  ...  We present an integrated approach for perception and control for an autonomous vehicle and demonstrate this approach in a high-fidelity urban driving simulator.  ...  Our key contributions can be summarized as: • We demonstrate a model-based reinforcement learning methodology for autonomous driving in which a model of complex driving environment is built and utilized  ... 
arXiv:2003.08300v1 fatcat:x6v4cvx65rbvfpqayzwplsathe

Behavioral decision-making for urban autonomous driving in the presence of pedestrians using Deep Recurrent Q-Network [article]

Niranjan Deshpande
2020 arXiv   pre-print
In this work, a deep reinforcement learning based decision-making approach for high-level driving behavior is proposed for urban environments in the presence of pedestrians.  ...  However, current research in the field of reinforcement learning for autonomous driving is mainly focused on highway setup with little to no emphasis on urban environments.  ...  ACKNOWLEDGMENT This work was funded under project CAMPUS (Connected Automated Mobilty Platform for Urban Sustainability) sponsored by Programme d'Investissements d'Avenir (PIA) of french Agence de l'Environnement  ... 
arXiv:2010.13407v1 fatcat:ytgezd3alzddbpl4ws3cgbddje

A Deep Reinforcement-Learning-based Driving Policy for Autonomous Road Vehicles [article]

Konstantinos Makantasis, Maria Kontorinaki, Ioannis Nikolos
2019 arXiv   pre-print
On the contrary, we propose the development of a driving policy based on reinforcement learning.  ...  To the best of our knowledge, this is one of the first approaches that propose a reinforcement learning driving policy for mixed driving environments.  ...  Acknowledgement This research is implemented through and has been financed by the Operational Program "Human Resources Development, Education and Lifelong Learning" and is co-financed by the European Union  ... 
arXiv:1907.05246v1 fatcat:zravjvkzybdpnipyflag5u434u

Policy-Gradient and Actor-Critic Based State Representation Learning for Safe Driving of Autonomous Vehicles

Abhishek Gupta, Ahmed Shaharyar Khwaja, Alagan Anpalagan, Ling Guan, Bala Venkatesh
2020 Sensors  
In this paper, we propose an environment perception framework for autonomous driving using state representation learning (SRL).  ...  is awarded for favourable actions.  ...  Figure 1 . 1 A comparison of data-driven and deep reinforcement learning based approaches to autonomous driving [31, 32] .  ... 
doi:10.3390/s20215991 pmid:33105863 pmcid:PMC7660054 fatcat:zpzg6u7dnzhopatydej35bh62m

A Deep Reinforcement-Learning-based Driving Policy for Autonomous Road Vehicles

Konstantinos Makantasis, Maria Kontorinaki, Ioannis Nikolos
2019 IET Intelligent Transport Systems  
On the contrary, this work proposes the development of a driving policy based on reinforcement learning.  ...  To the best of the authors' knowledge, this is one of the first approaches that propose a reinforcement learning driving policy for mixed driving environments.  ...  Acknowledgment This research was implemented through and has been financed by the Operational Program 'Human Resources Development, Education and Lifelong Learning' and is co-financed by the European Union  ... 
doi:10.1049/iet-its.2019.0249 fatcat:hr2dunbncfh53hgqh5i32lgrd4

Towards Safe, Explainable, and Regulated Autonomous Driving [article]

Shahin Atakishiyev, Mohammad Salameh, Hengshuai Yao, Randy Goebel
2022 arXiv   pre-print
However, as demonstrated by recent traffic accidents, autonomous driving technology is not mature for safe deployment.  ...  (AI), especially in the applications of deep learning and reinforcement learning.  ...  directions on explainable AI-based autonomous driving.  ... 
arXiv:2111.10518v3 fatcat:topadg7bp5enhflk7yqm6j27ga

Deep Reinforcement Learning for Autonomous Driving: A Survey [article]

B Ravi Kiran, Ibrahim Sobh, Victor Talpaert, Patrick Mannion, Ahmad A. Al Sallab, Senthil Yogamani, Patrick Pérez
2021 arXiv   pre-print
in real world deployment of autonomous driving agents.  ...  With the development of deep representation learning, the domain of reinforcement learning (RL) has become a powerful learning framework now capable of learning complex policies in high dimensional environments  ...  [75] provided a comprehensive review of the different state and action representations which are used in autonomous driving research.  ... 
arXiv:2002.00444v2 fatcat:axj3ohhjwzdrxp6dgpfqvctv2i

Model-free Deep Reinforcement Learning for Urban Autonomous Driving [article]

Jianyu Chen, Bodi Yuan, Masayoshi Tomizuka
2019 arXiv   pre-print
In this paper, we propose a framework to enable model-free deep reinforcement learning in challenging urban autonomous driving scenarios.  ...  Urban autonomous driving decision making is challenging due to complex road geometry and multi-agent interactions.  ...  Recent advances in machine learning enables the possibility for learning based approaches for autonomous driving decision making.  ... 
arXiv:1904.09503v2 fatcat:wxfrqwuowvhihh3h75mixlcyie

Lane Change Decision-making through Deep Reinforcement Learning with Rule-based Constraints [article]

Junjie Wang, Qichao Zhang, Dongbin Zhao, Yaran Chen
2019 arXiv   pre-print
Combined with the rule-based constraints, a Deep Q-Network (DQN) based method is applied for autonomous driving lane change decision-making task in this study.  ...  With the setting of our state representation and reward function, the trained agent is able to take appropriate actions in a real-world-like simulator.  ...  As a core technology of AI, learning-based approaches can provide more advanced and safe decision-making algorithms for autonomous driving.  ... 
arXiv:1904.00231v2 fatcat:nk2n7roqaba4dn4apaj67dc6xm

Safe Reinforcement Learning with Mixture Density Network: A Case Study in Autonomous Highway Driving [article]

Ali Baheri
2020 arXiv   pre-print
On the other hand, the learning-based safety module is a data-driven safety rule that learns safety patterns from driving data.  ...  This paper presents a safe reinforcement learning system for automated driving that benefits from multimodal future trajectory predictions.  ...  To collect data for MD-RNN training, we train an RL agent without a learning-based safety module and collect a long history of states and corresponding action that builds our driving data.  ... 
arXiv:2007.01698v3 fatcat:g6zxygzyenf7rjdlbr26x2fnbu

Practical Issues of Action-conditioned Next Image Prediction [article]

Donglai Zhu, Hao Chen, Hengshuai Yao, Masoud Nosrati, Peyman Yadmellat, Yunfei Zhang
2018 arXiv   pre-print
We provide the first comparison of two recent popular models, especially for image prediction on cars.  ...  The problem of action-conditioned image prediction is to predict the expected next frame given the current camera frame the robot observes and an action selected by the robot.  ...  Besides visualization, building action models is essential to bring reinforcement learning to autonomous driving.  ... 
arXiv:1802.02975v1 fatcat:eh5554bn6zgsxklih5m73p6xsu

Towards Learning Generalizable Driving Policies from Restricted Latent Representations [article]

Behrad Toghi, Rodolfo Valiente, Ramtin Pedarsani, Yaser P. Fallah
2022 arXiv   pre-print
It goes without saying that although scenario-specific driving policies for autonomous driving are promising and can improve transportation safety and efficiency, they are clearly not a universal scalable  ...  This latent space is then employed as the input to a Q-learning module to learn generalizable driving policies.  ...  The learned latent representation is used for learning driving policies via Q-learning.  ... 
arXiv:2111.03688v2 fatcat:giliu7j5zzc6riblt4zcs6we6u

Autonomous Driving in Reality with Reinforcement Learning and Image Translation [article]

Nayun Xu, Bowen Tan, Bingyu Kong
2019 arXiv   pre-print
Supervised learning is widely used in training autonomous driving vehicle. However, it is trained with large amount of supervised labeled data.  ...  Reinforcement learning can be trained without abundant labeled data, but we cannot train it in reality because it would involve many unpredictable accidents.  ...  Related Work Supervised Learning for Autonomous Driving Supervised learning has been used in autonomous driving for decades.  ... 
arXiv:1801.05299v2 fatcat:rsf4qxdb3ffehj7kprt2eybjhm

Learning by Watching [article]

Jimuyang Zhang, Eshed Ohn-Bar
2021 arXiv   pre-print
In contrast, existing techniques for learning to drive preclude such a possibility as they assume direct access to an instrumented ego-vehicle with fully known observations and expert driver actions.  ...  Motivated by this key insight, we propose the Learning by Watching (LbW) framework which enables learning a driving policy without requiring full knowledge of neither the state nor expert actions.  ...  Imitation Learning for Autonomous Driving: Over 30 years ago, Pomerleau [55] developed ALVINN, a neural network-based approach for learning to imitate a driver of an ego-vehicle.  ... 
arXiv:2106.05966v1 fatcat:eabm4ktwpje4toirfhd5krtv6q
« Previous Showing results 1 — 15 out of 79,960 results