The DOI matches its Crossref record; 218 of the 230 references it cites were found in a registry or an index.
A Survey of Deep Learning Applications to Autonomous Vehicle Control
Open accessAbstract
Designing a controller for autonomous vehicles capable of providing adequate performance in all driving scenarios is challenging due to the highly complex environment and inability to test the system in the wide variety of scenarios which it may encounter after deployment. However, deep learning methods have shown great promise in not only providing excellent performance for complex and non-linear control problems, but also in generalising previously learned rules to new scenarios. For these reasons, the use of deep learning for vehicle control is becoming increasingly popular. Although important advancements have been achieved in this field, these works have not been fully summarised. This paper surveys a wide range of research works reported in the literature which aim to control a vehicle through deep learning methods. Although there exists overlap between control and perception, the focus of this paper is on vehicle control, rather than the wider perception problem which includes tasks such as semantic segmentation and object detection. The paper identifies the strengths and limitations of available deep learning methods through comparative analysis and discusses the research challenges in terms of computation, architecture selection, goal specification, generalisation, verification and validation, as well as safety. Overall, this survey brings timely and topical information to a rapidly evolving field relevant to intelligent transportation systems.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This includes historical signals and inappropriately proxied measures that make ML systems vulnerable to reproducing past discrimination against under-represented groups (e.g. in contexts such as job hiring and criminal justice), or propagating abusive content [46,67].”
From A Survey of Data Quality Requirements That Matter in ML Development Pipelines (2023) · cites this paper as [46]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This includes historical signals and inappropriately proxied measures that make ML systems vulnerable to reproducing past discrimination against under-represented groups (e.g. in contexts such as job hiring and criminal justice), or propagating abusive content [46,67].”
From A Survey of Data Quality Requirements That Matter in ML Development Pipelines (2023) · cites this paper as [46]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“in fields such as image classification and speech recognition [24]–[26].”
BackgroundAdam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 2×
“Then, the neural network was trained with three different optimisation algorithms to update the network weights, namely Stochastic Gradient Descent (SGD) [90], Adam [91], and Nesterov’s Accelerated Gradient (NAG) [92].”
MethodLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
(2014) · Cited by 33,419
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357 · cited 1×
“deep learning enables developers to describe the desired behaviour and teach the system to perform well and generalise to new environments through learning [43]–[47].”
BackgroundReinforcement Learning: An Introduction
Sutton, R. S., & Barto, A. (2005) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2004.842673 · Cited by 25,268
Deep learning in neural networks: an overview
Schmidhuber, J. (2015) · Neural Networks · doi:10.1016/j.neunet.2014.09.003 · Cited by 18,570
Neural networks for pattern recognition
Bishop, C. (1994) · Choice Reviews Online · doi:10.5860/choice.31-5500 · Cited by 18,537 · cited 1×
“work and to terminate training can result in overfitting to the validation set. Therefore an additional independent set, known as testing set, is required for the evaluation of the network performance [186]. The testing set is only used to test the final network to confirm its performance and generalisation capabilities.…”
MethodSequence to Sequence Learning with Neural Networks
Sutskever, I., Vinyals, O., & Le, Q. V. (2014) · arXiv · doi:10.48550/arxiv.1409.3215 · Cited by 17,766 · cited 1×
“in fields such as image classification and speech recognition [24]–[26].”
BackgroundAre we ready for autonomous driving? The KITTI vision benchmark suite
Geiger, A., Lenz, P., & Urtasun, R. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248074 · Cited by 14,897
Bellman, R. E. (2010) · doi:10.1515/9781400835386 · Cited by 14,588
(2016) · Cited by 11,679
Continuous control with deep reinforcement learning
Lillicrap, T. P., Hunt, J. J., Pritzel, A., Heess, N., Erez, T., Tassa, Y., Silver, D., & Wierstra, D. (2015) · arXiv · doi:10.48550/arxiv.1509.02971 · Cited by 10,782 · cited 1×
“This would require an iterative process at every time step for continuous action spaces [124].”
Explaining and Harnessing Adversarial Examples
Goodfellow, I. J., Shlens, J., & Szegedy, C. (2014) · arXiv · doi:10.48550/arxiv.1412.6572 · Cited by 10,381 · cited 1×
“It has been shown that visual classification DNN systems are vulnerable to adversarial examples, which are perturbed images that cause the DNNs to misclassify them with high confidence [216]–[219], including misclassification of traffic signs [220].”
BackgroundHinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A.-R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T., et al. (2012) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2205597 · Cited by 10,260
Vision meets robotics: The KITTI dataset
Geiger, A., Lenz, P., Stiller, C., & Urtasun, R. (2013) · The International Journal of Robotics Research · doi:10.1177/0278364913491297 · Cited by 9,913 · cited 1×
“Perhaps the most well known data set for autonomous driving is the KITTI benchmark suite [70], [71], which includes multiple data sets for evaluation of stereo vision, optical flow, scene flow, simultaneous localisation and mapping, object detection and tracking, road detection and”
BackgroundWatkins, C. J. C. H., & Dayan, P. (1992) · Machine Learning · doi:10.1007/bf00992698 · Cited by 9,083
Reinforcement Learning: A Survey
Kaelbling, L. P., Littman, M. L., & Moore, A. W. (1996) · Journal of Artificial Intelligence Research · doi:10.1613/jair.301 · Cited by 8,972 · cited 1×
“However, the main drawback for reinforcement learning is the time-intensive training [50], [114].”
BackgroundIntriguing properties of neural networks
Szegedy, C., Zaremba, W., Sutskever, I., Bruna, J., Erhan, D., Goodfellow, I., & Fergus, R. (2013) · arXiv · doi:10.48550/arxiv.1312.6199 · Cited by 8,143 · cited 1×
“It has been shown that visual classification DNN systems are vulnerable to adversarial examples, which are perturbed images that cause the DNNs to misclassify them with high confidence [216]–[219], including misclassification of traffic signs [220].”
Background(2012) · Cited by 8,009
Asynchronous Methods for Deep Reinforcement Learning
Mnih, V., Badia, A. P., Mirza, M., Graves, A., Lillicrap, T. P., Harley, T., Silver, D., & Kavukcuoglu, K. (2016) · arXiv · doi:10.48550/arxiv.1602.01783 · Cited by 7,132 · cited 2×
“The third class, actor-critic algorithms (e.g. A3C [67]), are hybrid methods which combine the use of a value function with a parametrised policy function.”
MethodPractical Bayesian Optimization of Machine Learning Algorithms
Snoek, J., Larochelle, H., & Adams, R. P. (2012) · arXiv · doi:10.48550/arxiv.1206.2944 · Cited by 6,691
Gaussian Processes in Machine Learning
Rasmussen, C. E. (2004) · Lecture notes in computer science · doi:10.1007/978-3-540-28650-9_4 · Cited by 6,522
Improving neural networks by preventing co-adaptation of feature detectors
Hinton, G. E., Srivastava, N., Krizhevsky, A., Sutskever, I., & Salakhutdinov, R. R. (2012) · arXiv · doi:10.48550/arxiv.1207.0580 · Cited by 6,382
Large-Scale Machine Learning with Stochastic Gradient Descent
Bottou, L. (2010) · Proceedings of COMPSTAT'2010 · doi:10.1007/978-3-7908-2604-3_16 · Cited by 5,795 · cited 1×
“Then, the neural network was trained with three different optimisation algorithms to update the network weights, namely Stochastic Gradient Descent (SGD) [90], Adam [91], and Nesterov’s Accelerated Gradient (NAG) [92].”
MethodAdversarial examples in the physical world
Kurakin, A., Goodfellow, I. J., & Bengio, S. (2018) · Artificial Intelligence Safety and Security · doi:10.1201/9781351251389-8 · Cited by 5,375
DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks
Moosavi-Dezfooli, S.-M., Fawzi, A., & Frossard, P. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.282 · Cited by 5,317
(1999) · Cited by 5,147
Deep Reinforcement Learning: A Brief Survey
Arulkumaran, K., Deisenroth, M. P., Brundage, M., & Bharath, A. A. (2017) · IEEE Signal Processing Magazine · doi:10.1109/msp.2017.2743240 · Cited by 4,733 · cited 2×
“Therefore, intrinsic motivation is implemented in the system, encouraging the agent to learn about its environment, whilst exploitation can be taken advantage of in states which have already been explored adequately [51], [117]–[119].”
Background(2011) · Cited by 3,939
CARLA: An open urban driving simulator
Dosovitskiy, A., Ros, G., Codevilla, F., Lopez, A., & Koltun, V. (2017) · arXiv · doi:10.48550/arxiv.1711.03938 · Cited by 3,718 · cited 1×
“The resulting network was initially tested in CARLA [139] simulation, followed by real-world testing on a 1/5-scale car.”
MethodDesign and use paradigms for gazebo, an open-source multi-robot simulator
Koenig, N., & Howard, A. (2005) · 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566) · doi:10.1109/iros.2004.1389727 · Cited by 3,638 · cited 2×
“The technique was first tested in Robot Operation System (ROS) Gazebo [127] simulations,”
MethodMnasNet: Platform-Aware Neural Architecture Search for Mobile
Tan, M., Chen, B., Pang, R., Vasudevan, V., Sandler, M., Howard, A., & Le, Q. V. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00293 · Cited by 3,416
Deterministic Policy Gradient Algorithms
Silver, D., Lever, G., Heess, N., Degris, T., Wierstra, D., & Riedmiller, M. (2014) · HAL · Cited by 3,275
End to End Learning for Self-Driving Cars
Bojarski, M., Del Testa, D., Dworakowski, D., Firner, B., Flepp, B., Goyal, P., Jackel, L. D., Monfort, M., Muller, U., Zhang, J., et al. (2016) · arXiv · doi:10.48550/arxiv.1604.07316 · Cited by 3,198 · cited 2×
“The neural networks utilised in the aforementioned early works are significantly smaller when compared to what is feasible with today’s technology [87].”
BackgroundMethodReinforcement learning in robotics: A survey
Kober, J., Bagnell, J. A., & Peters, J. (2013) · The International Journal of Robotics Research · doi:10.1177/0278364913495721 · Cited by 3,160 · cited 2×
“However, as the number of dimensions grows, the computational complexity grows exponentially [152], this is known as the Curse of Dimensionality [153].”
BackgroundNeural Networks and Deep Learning
Aggarwal, C. C. (2018) · Springer eBooks · doi:10.1007/978-3-319-94463-0 · Cited by 3,154
Domain randomization for transferring deep neural networks from simulation to the real world
Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., & Abbeel, P. (2017) · 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8202133 · Cited by 3,124 · cited 1×
“policies from simulation to the real world [199]–[202].”
BackgroundApprenticeship learning via inverse reinforcement learning
Abbeel, P., & Ng, A. Y. (2004) · International Conference on Machine Learning · doi:10.1145/1015330.1015430 · Cited by 2,855 · cited 1×
“Abbeel & Ng [131] showed that when IRL was applied to a problem where the agent learned by observing an expert, the agent performed as well as the expert when evaluated with respect to the reward function used by the expert, even if the reward function derived from observations was not the expert’s true reward functio…”
BackgroundA Survey of Motion Planning and Control Techniques for Self-Driving Urban Vehicles
Paden, B. A., Čáp, M., Yong, S. Z., Yershov, D. S., & Frazzoli, E. (2016) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2016.2578706 · Cited by 2,724
End-to-End Training of Deep Visuomotor PoliciesPreprint
Levine, S., Finn, C., Darrell, T., & Abbeel, P. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1504.00702 · Cited by 2,641 · cited 2×
“problems in complex and dynamic environments [40]–[42].”
BackgroundTorralba, A., & Efros, A. A. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995347 · Cited by 2,440 · cited 2×
“Therefore, data sets should be validated to understand if they contain potentially harmful biases or patterns that could lead to undesirable behaviour of the learned control policy [56].”
BackgroundAsymptotically efficient adaptive allocation rules
Lai, T. L., & Robbins, H. (1985) · Advances in Applied Mathematics · doi:10.1016/0196-8858(85)90002-8 · Cited by 2,406
Practical Recommendations for Gradient-Based Training of Deep Architectures
Bengio, Y. (2012) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-35289-8_26 · Cited by 2,259 · cited 5×
“Random Search often finds a good set of parameters faster than a Grid Search by sampling the chosen interval randomly [162], [163].”
BackgroundInfluentialStanley: The robot that won the DARPA Grand Challenge
Thrun, S., Montemerlo, M., Dahlkamp, H., Stavens, D., Aron, A., Diebel, J., Fong, P., Gale, J., Halpenny, M., Hoffmann, G., et al. (2006) · Journal of Field Robotics · doi:10.1002/rob.20147 · Cited by 2,237
BinaryConnect: Training Deep Neural Networks with binary weights during propagations
Courbariaux, M., Bengio, Y., & David, J.-P. (2015) · arXiv · doi:10.48550/arxiv.1511.00363 · Cited by 2,180 · cited 1×
“Similarly, imposing constraints on the network weights through weight clipping has also been shown to improve robustness [188], [189].”
BackgroundHigh-Dimensional Continuous Control Using Generalized Advantage Estimation
Schulman, J., Moritz, P., Levine, S., Jordan, M., & Abbeel, P. (2015) · arXiv · doi:10.48550/arxiv.1506.02438 · Cited by 2,103
A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning
Ross, S., Gordon, G. J., & Bagnell, J. A. (2018) · Carnegie Mellon University · doi:10.1184/r1/6550949 · Cited by 1,941
Neural networks for pattern recognition
Marshall, J. A. (1995) · Neural Networks · doi:10.1016/0893-6080(95)90002-0 · Cited by 1,888
Maximum Entropy Inverse Reinforcement Learning
Ziebart, B. D., Maas, A. L., Bagnell, J. A., & Dey, A. K. (2018) · doi:10.1184/r1/6555512 · Cited by 1,868
Can we open the black box of AI?
Castelvecchi, D. (2016) · Nature · doi:10.1038/538020a · Cited by 1,773 · cited 1×
“As the neural networks become more complex, the solutions they provide and how they come to those solutions becomes increasingly difficult to interpret [204].”
BackgroundNeural Architecture Search: A Survey
Elsken, T., Metzen, J. H., & Hutter, F. (2018) · arXiv · doi:10.48550/arxiv.1808.05377 · Cited by 1,763
Policy Invariance Under Reward Transformations: Theory and Application to Reward Shaping
Ng, A. Y., Harada, D., & Russell, S. (1999) · Cited by 1,744
Adversarial Examples: Attacks and Defenses for Deep Learning
Yuan, X., He, P., Zhu, Q., & Li, X. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2886017 · Cited by 1,733
1 year, 1000 km: The Oxford RobotCar dataset
Maddern, W., Pascoe, G., Linegar, C., & Newman, P. (2017) · The International Journal of Robotics Research · doi:10.1177/0278364916679498 · Cited by 1,632 · cited 2×
“Other useful data sets include the Waymo Open [72], Oxford Robotcar [73], ApolloScape [74], Udacity [75], ETH Pedestrian [76], and Caltech Pedestrian [77] data sets.”
MethodFeature selection, L 1 vs. L 2 regularization, and rotational invariance
Ng, A. Y. (2004) · Twenty-first international conference on Machine learning - ICML '04 · doi:10.1145/1015330.1015435 · Cited by 1,572 · cited 1×
“e network reduce the effect of individual inputs on its behaviour, which means that the effect of local noise is reduced and the network is more likely to learn trends across the whole data set [49], [187]. Similarly, imposing constraints on the network weights through weight clipping has also been shown to improve ro…”
BackgroundStable adaptive neural control scheme for nonlinear systems
Polycarpou, M. (1996) · IEEE Transactions on Automatic Control · doi:10.1109/9.486648 · Cited by 1,539
ALVINN, an autonomous land vehicle in a neural network
Pomerleau, D. (2018) · doi:10.1184/r1/6603146 · Cited by 1,369
Pedestrian detection: A benchmark
Dollár, P., Wojek, C., Schiele, B., & Perona, P. (2009) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2009.5206631 · Cited by 1,351
FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search
Wu, B., Keutzer, K., Dai, X., Zhang, P., Wang, Y., Sun, F., Wu, Y., Tian, Y., Vajda, P., & Jia, Y. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.01099 · Cited by 1,349
LeCun, Y., Bottou, L., Orr, G., & Müller, K. (1998) · neural information processing systems · Cited by 1,333
Concrete Problems in AI Safety
Amodei, D., Olah, C., Steinhardt, J., Christiano, P., Schulman, J., & Mané, D. (2016) · arXiv · doi:10.48550/arxiv.1606.06565 · Cited by 1,314 · cited 2×
“[179] in the form of adversarial reward functions, model look-ahead, reward capping, multiple reward functions, and trip wires.”
BackgroundWang, D., & Huang, J. (2005) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2004.839354 · Cited by 1,267
Deep Machine Learning - A New Frontier in Artificial Intelligence Research [Research Frontier]
Arel, I., Rose, D. C., & Karnowski, T. P. (2010) · IEEE Computational Intelligence Magazine · doi:10.1109/mci.2010.938364 · Cited by 1,252
The Tradeoffs of Large-Scale Learning
Bottou, L., & Bousquet, O. (2011) · The MIT Press eBooks · doi:10.7551/mitpress/8996.003.0015 · Cited by 1,201
PILCO: A Model-Based and Data-Efficient Approach to Policy Search
Deisenroth, M. P., & Rasmussen, C. E. (2011) · Cambridge University Engineering Department Publications Database · Cited by 1,136
Synthesizing Robust Adversarial ExamplesPreprint
Athalye, A., Engstrom, L., Ilyas, A., & Kwok, K. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.07397 · Cited by 1,133 · cited 1×
“to printed adversarial examples in the real world [221] and even to 3D-printed physical adversarial examples [222], which suggests they are a threat to DNN applications in the real world.”
BackgroundA Survey of Actor-Critic Reinforcement Learning: Standard and Natural Policy Gradients
Grondman, I., Buşoniu, L., Lopes, G. A. D., & Babuška, R. (2012) · IEEE Transactions on Systems Man and Cybernetics Part C (Applications and Reviews) · doi:10.1109/tsmcc.2012.2218595 · Cited by 1,091
Gaussian Networks for Direct Adaptive Control
Sanner, R. M., & Slotine, J.-J. E. (1991) · Proceedings of the ... American Control Conference/Proceedings of the American Control Conference · doi:10.23919/acc.1991.4791778 · Cited by 1,079
Visual Place Recognition: A Survey
Lowry, S., Sünderhauf, N., Newman, P., Leonard, J. J., Cox, D., Corke, P., & Milford, M. J. (2016) · IEEE Transactions on Robotics · doi:10.1109/tro.2015.2496823 · Cited by 1,079 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundEvolution Strategies as a Scalable Alternative to Reinforcement Learning
Salimans, T., Ho, J., Chen, X., Sidor, S., & Sutskever, I. (2017) · arXiv · doi:10.48550/arxiv.1703.03864 · Cited by 1,027 · cited 1×
“lelisable [156], or removing unnecessary data from the training”
BackgroundPlanning and Decision-Making for Autonomous Vehicles
Schwarting, W., Alonso-Mora, J., & Rus, D. (2018) · Annual Review of Control, Robotics, and Autonomous Systems · doi:10.1146/annurev-control-060117-105157 · Cited by 972 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundLeCun, Y., Bottou, L., Orr, G., & Müller, K. R. (1998) · Lecture notes in computer science · doi:10.1007/3-540-49430-8_2 · Cited by 920
Safety Verification of Deep Neural Networks
Huang, X., Kwiatkowska, M., Wang, S., & Wu, M. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-63387-9_1 · Cited by 903
Deep Q-learning From Demonstrations
Hester, T., Vecerik, M., Pietquin, O., Lanctot, M., Schaul, T., Piot, B., Horgan, D., Quan, J., Sendonaris, A., Osband, I., et al. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11757 · Cited by 889
Generalization and network design strategies
LeCun, Y. (1989) · Cited by 885
Autonomous vehicle perception: The technology of today and tomorrow
Van Brummelen, J., O’Brien, M., Gruyer, D., & Najjaran, H. (2018) · Transportation Research Part C: Emerging Technologies · doi:10.1016/j.trc.2018.02.012 · Cited by 869
Research advances in intelligent collision avoidance and adaptive cruise control
Vahidi, A., & Eskandarian, A. (2003) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2003.821292 · Cited by 848 · cited 1×
“control problem for a complex nonlinear system [97], [98] and therefore is poorly suited to control systems based on linear”
BackgroundOnce-for-All: Train One Network and Specialize it for Efficient Deployment
Cai, H., Gan, C., Wang, T., Zhang, Z., & Han, S. (2019) · arXiv · doi:10.48550/arxiv.1908.09791 · Cited by 845
Autonomous vehicles: challenges, opportunities, and future implications for transportation policies
Bagloee, S. A., Madjid Tavana, D., Asadi, M., & Oliver, T. (2016) · 西南交通大学学报：英文版 · doi:10.1007/s40534-016-0117-3 · Cited by 828
Riedmiller, M. (2005) · Lecture notes in computer science · doi:10.1007/11564096_32 · Cited by 824 · cited 3×
“Moreover, during 50 tests on a competition track, the proposed approach completed the track 49 times, compared to only 33 with NFQ. Additionally, DQFE performed better in terms of mean distance from centre of the track.”
MethodInfluential(2015) · Cited by 809
Kuutti, S., Fallah, S., Katsaros, K., Dianati, M., Mccullough, F., & Mouzakitis, A. (2018) · IEEE Internet of Things Journal · doi:10.1109/jiot.2018.2812300 · Cited by 809 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundSelf-citationContinuous Deep Q-Learning with Model-based AccelerationPreprint
Gu, S., Lillicrap, T., Sutskever, I., & Levine, S. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.00748 · Cited by 770 · cited 1×
“Other solutions include using multiple learners to reduce learning time [154], [155], evolution strategies which are highly parallelisable [156], or removing unnecessary data from the training and system input data [157].”
BackgroundDeep Reinforcement Learning: An Overview
Li, Y. (2017) · arXiv · doi:10.48550/arxiv.1701.07274 · Cited by 762 · cited 1×
“manuscript, we refer the interested reader to the insightful texts on this topic in [46]–[52].”
BackgroundBuehler, M., Iagnemma, K. D., & Singh, S. (2009) · Springer tracts in advanced robotics · doi:10.1007/978-3-642-03991-1 · Cited by 758 · cited 1×
“Since then, projects such as DARPA Grand Challenges [15], [16] have continued to drive forward research in autonomous vehicles.”
BackgroundChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst
Bansal, M., Krizhevsky, A., & Ogale, A. S. (2019) · Robotics: Science and Systems · doi:10.15607/rss.2019.xv.031 · Cited by 743
(2016) · neural information processing systems · Cited by 735
Sample Efficient Actor-Critic with Experience ReplayPreprint
Wang, Z., Bapst, V., Heess, N., Mnih, V., Munos, R., Kavukcuoglu, K., & de Freitas, N. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01224 · Cited by 729 · cited 1×
“The downside of reinforcement learning is its low sample efficiency [58], which means converging to an optimal policy can be slow, thereby requiring time-intensive simulations or costly real-world training [50].”
BackgroundKonda, V. R., & Tsitsiklis, J. N. (2003) · SIAM Journal on Control and Optimization · doi:10.1137/s0363012901385691 · Cited by 721
A mobile vision system for robust multi-person tracking
Ess, A., Leibe, B., Schindler, K., & Van Gool, L. (2008) · 2008 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2008.4587581 · Cited by 721 · cited 1×
“Other useful data sets include the Waymo Open [72], Oxford Robotcar [73], ApolloScape [74], Udacity [75], ETH Pedestrian [76], and Caltech Pedestrian [77] data sets.”
MethodA Differential Equation for Modeling Nesterov's Accelerated Gradient Method: Theory and InsightsPreprint
Su, W., Boyd, S., & Candès, E. J. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1503.01243 · Cited by 713 · cited 3×
“Then, the neural network was trained with three different optimisation algorithms to update the network weights, namely Stochastic Gradient Descent (SGD) [90], Adam [91], and Nesterov’s Accelerated Gradient (NAG) [92].”
BackgroundDEX: Deep EXpectation of Apparent Age from a Single Image
Rothe, R., Timofte, R., & Gool, L. V. (2015) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2015.41 · Cited by 711 · cited 1×
“angles based on image inputs, and then compared it to a simple CNN architecture used in [94].”
MethodComputer Vision for Autonomous Vehicles
Janai, J., Güney, F., Behl, A., & Geiger, A. (2020) · Foundations and Trends® in Computer Graphics and Vision · doi:10.1561/0600000079 · Cited by 707
Intention to use a fully automated car: Attitudes and a priori acceptability
Payre, W., Cestac, J., & Delhomme, P. (2014) · Transportation Research Part F Traffic Psychology and Behaviour · doi:10.1016/j.trf.2014.04.009 · Cited by 696 · cited 1×
“autonomous vehicles in terms of better fuel economy [9], [10], reduced pollution, car sharing [11], increased productivity, and”
BackgroundAdaptive neural network control for strict-feedback nonlinear systems using backstepping design
Zhang, T., Ge, S. S., & Hang, C. C. (2000) · Automatica · doi:10.1016/s0005-1098(00)00116-3 · Cited by 682
The ApolloScape Dataset for Autonomous Driving
Huang, X., Cheng, X., Geng, Q., Cao, B., Zhou, D., Wang, P., Lin, Y., & Yang, R. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2018.00141 · Cited by 679
Heuristic approaches in robot path planning: A survey
Mac, T. T., Copot, C., Tran, D. T., & KEYSER, R. D. (2016) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2016.08.001 · Cited by 643
Ratliff, N., Bagnell, J. A., & Zinkevich, M. (2006) · International Conference on Machine Learning · doi:10.1145/1143844.1143936 · Cited by 636 · cited 2×
“[21] used an IRL algorithm based on Maximum Margin Planning [132] which was shown to be effective in a demonstration of an autonomous vehicle in unstructured terrain.”
MethodThe ApolloScape Open Dataset for Autonomous Driving and Its Application
Huang, X., Wang, P., Cheng, X., Zhou, D., Geng, Q., & Yang, R. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2926463 · Cited by 615 · cited 1×
“Other useful data sets include the Waymo Open [72], Oxford Robotcar [73], ApolloScape [74], Udacity [75], ETH Pedestrian [76], and Caltech Pedestrian [77] data sets.”
MethodTen Years of Pedestrian Detection, What Have We Learned?
Benenson, R., Omran, M., Hosang, J., & Schiele, B. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-16181-5_47 · Cited by 609
Exploring the Limitations of Behavior Cloning for Autonomous Driving
Codevilla, F., Santana, E., López, A. M., & Gaidon, A. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00942 · Cited by 608 · cited 3×
“However, as noted in [140], when large amounts of data are available, deeper architectures can reduce both bias and variance in training, resulting in more robust control policies.”
BackgroundMethodInfluentialRajamani, R., Tan, H.-S., Law, B., & Zhang, W.-B. (2000) · IEEE Transactions on Control Systems Technology · doi:10.1109/87.852914 · Cited by 581
An Application of Reinforcement Learning to Aerobatic Helicopter Flight
Abbeel, P., Coates, A., Quigley, M., & Ng, A. Y. (2007) · The MIT Press eBooks · doi:10.7551/mitpress/7503.003.0006 · Cited by 551 · cited 1×
“[136], who used IRL to operate an autonomous helicopter and had to manually tune the reward function for safety.”
MethodSafe, Multi-Agent, Reinforcement Learning for Autonomous Driving
Shalev-Shwartz, S., Shammah, S., & Shashua, A. (2016) · arXiv · doi:10.48550/arxiv.1610.03295 · Cited by 509
How Far are We from Solving Pedestrian Detection?
Zhang, S., Benenson, R., Omran, M., Hosang, J., & Schiele, B. (2016) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2016.141 · Cited by 481 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundA review of design and control of automated guided vehicle systems
Le–Anh, T., & de Koster, R. B. M. (2005) · European Journal of Operational Research · doi:10.1016/j.ejor.2005.01.036 · Cited by 476
Learning driving styles for autonomous vehicles from demonstration
Kuderer, M., Gulati, S., & Burgard, W. (2015) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2015.7139555 · Cited by 468 · cited 2×
“The downside of this approach is the time intensive hand-tuning of parameters [20] and the difficulty of”
BackgroundStable Function Approximation in Dynamic Programming
Gordon, G. J. (1995) · Elsevier eBooks · doi:10.1016/b978-1-55860-377-6.50040-2 · Cited by 458 · cited 1×
“The disadvantage of this approach is that there is no guarantee on the optimality of the learned policy [61], [62].”
BackgroundA Possibility for Implementing Curiosity and Boredom in Model-Building Neural Controllers
Schmidhuber, J. (1991) · The MIT Press eBooks · doi:10.7551/mitpress/3115.003.0030 · Cited by 455 · cited 1×
“Therefore, intrinsic motivation is implemented in the system, encouraging the agent to learn about its environment, whilst exploitation can be taken advantage of in states which have already been explored adequately [51], [117]–[119].”
BackgroundFeature-Based Methods for Large Scale Dynamic Programming
Tsitsiklis, J. N., & Roy, B. V. (1996) · Machine Learning · doi:10.1007/bf00114724 · Cited by 452 · cited 1×
“The disadvantage of this approach is that there is no guarantee on the optimality of the learned policy [61], [62].”
Background(2005) · neural information processing systems · Cited by 446
Synthesizing Robust Adversarial Examples
Athalye, A., Engstrom, L., Ilyas, A., & Kwok, K. (2018) · International Conference on Machine Learning · Cited by 432
The Robot That Won the DARPA Grand Challenge
Thrun, S. (2007) · Cited by 406
Kato, S., Tsugawa, S., Tokuda, K., Matsui, T., & Fujii, H. (2002) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2002.802929 · Cited by 402 · cited 1×
“Indeed, full-scale vehicle studies with multiple vehicles have typically been achieved through collaboration of government research projects with automotive manufacturers, such as Demo ’97 [192]–[194] or Demo 2000 [195].”
BackgroundOverview of Environment Perception for Intelligent Vehicles
Zhu, H., Yuen, K., Mihaylova, L. S., & Leung, H. (2017) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2017.2658662 · Cited by 386 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundSim-to-Real Robot Learning from Pixels with Progressive Nets
Rusu, A. A., Vecerik, M., Rothörl, T., Heess, N., Pascanu, R., & Hadsell, R. (2016) · arXiv · doi:10.48550/arxiv.1610.04286 · Cited by 385
Handbook of Intelligent Vehicles
Eskandarian, A. (2012) · Springer eBooks · doi:10.1007/978-0-85729-085-4 · Cited by 380 · cited 1×
“hicles on the road, issues such as traffic congestion, pollution, and road safety are becoming critical issues [3]. Autonomous vehicles have gained significant interest as solutions to these challenges [4]–[7]. For instance, 90% of all car accidents are estimated to be caused by human errors, while only 2% are caused by…”
BackgroundLand, M. F., & Fernald, R. D. (1992) · Annual Review of Neuroscience · doi:10.1146/annurev.ne.15.030192.000245 · Cited by 375
Adaptive neural network control for a class of uncertain nonlinear systems in pure-feedback form
Wang, D., & Huang, J. (2002) · Automatica · doi:10.1016/s0005-1098(02)00034-1 · Cited by 358
Cooperative Adaptive Cruise Control: A Reinforcement Learning Approach
Desjardins, C., & Chaib-draa, B. (2011) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2011.2157145 · Cited by 347 · cited 5×
“For example, Desjardins & Chaib-Draa [23] used the time headway derivative to reward the agent for actions that helped it move towards the ideal time headway state.”
BackgroundMethodInfluentialLearning agents for uncertain environments (extended abstract)
Russell, S. (1998) · Conference on Learning Theory · doi:10.1145/279943.279964 · Cited by 344 · cited 1×
“function from expert demonstrations of the task [180]–[183].”
BackgroundAutonomous Ground Vehicles—Concepts and a Path to the Future
Luettel, T., Himmelsbach, M., & Wuensche, H. (2012) · Proceedings of the IEEE · doi:10.1109/jproc.2012.2189803 · Cited by 339 · cited 1×
“autonomous vehicles in terms of better fuel economy [9], [10], reduced pollution, car sharing [11], increased productivity, and”
BackgroundGlaser, S., Vanholme, B., Mammar, S., Gruyer, D., & Nouvelière, L. (2010) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2010.2046037 · Cited by 336 · cited 1×
“Also, the safety based controller uses an Artificial Potential Field method [215] which models any obstacles with a repulsive force to steer the vehicle away from them.”
MethodVirtual to Real Reinforcement Learning for Autonomous Driving
Pan, X., You, Y., Wang, Z., & Lu, C. (2017) · Procedings of the British Machine Vision Conference 2017 · doi:10.5244/c.31.11 · Cited by 326 · cited 1×
“ferring driving policies from simulation to the real world easier [143], [144].”
BackgroundVirtual to Real Reinforcement Learning for Autonomous Driving
Pan, X., You, Y., Wang, Z., & Lu, C. (2017) · arXiv · doi:10.48550/arxiv.1704.03952 · Cited by 326 · cited 1×
“ferring driving policies from simulation to the real world easier [143], [144].”
BackgroundA Reinforcement Learning Based Approach for Automated Lane Change Maneuvers
Wang, P., Chan, C., & Fortelle, A. D. L. (2018) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2018.8500556 · Cited by 319
Moon, S., Moon, I., & Yi, K. (2008) · Control Engineering Practice · doi:10.1016/j.conengprac.2008.09.006 · Cited by 317 · cited 1×
“control problem for a complex nonlinear system [97], [98] and therefore is poorly suited to control systems based on linear”
BackgroundLearning Hand-Eye Coordination for Robotic Grasping with Large-Scale Data Collection
Levine, S., Pástor, P., Krizhevsky, A., & Quillen, D. (2017) · Springer proceedings in advanced robotics · doi:10.1007/978-3-319-50115-4_16 · Cited by 282 · cited 1×
“However, neural networks have shown great potential for optimising nonlinear, high-dimensional control systems [40], [41], [101]–[106].”
BackgroundKernel-Based Least Squares Policy Iteration for Reinforcement Learning
Xu, X., Hu, D., & Lu, X. (2007) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2007.899161 · Cited by 272 · cited 1×
“trollers trained with reinforcement learning [109], [110].”
Automated On-Ramp Merging System for Congested Traffic Situations
Milanés, V., Godoy, J., Villagrá, J., & Perez, J. (2010) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2010.2096812 · Cited by 257
Trajectory planning and tracking for autonomous overtaking: State-of-the-art and future prospects
Dixit, S., Fallah, S., Montanaro, U., Dianati, M., Stevens, A., Mccullough, F., & Mouzakitis, A. (2018) · Annual Reviews in Control · doi:10.1016/j.arcontrol.2018.02.001 · Cited by 257 · cited 1×
“in autonomous vehicle applications, including planning and decision making [27]–[31], perception [32]–[36], as well as mapping and localisation [37]–[39].”
BackgroundSelf-citationAgile Autonomous Driving using End-to-End Deep Imitation Learning
Pan, Y., Cheng, C.-A., Saigol, K., Lee, K., Yan, X., Theodorou, E., & Boots, B. (2018) · Robotics: Science and Systems XIV · doi:10.15607/rss.2018.xiv.056 · Cited by 257 · cited 1×
“[126] used DAgger-like imitation learning to learn to drive at high speeds autonomously, with continuous actions for both steering and acceleration.”
MethodAutonomous High Speed Road Vehicle Guidance by Computer Vision 1
Dickmanns, E. D., & Zapp, A. (1987) · IFAC Proceedings Volumes · doi:10.1016/s1474-6670(17)55320-3 · Cited by 252 · cited 1×
“in structured environments [13] and the University of Bundeswehr Munich for highway driving [14].”
BackgroundThrun, S. (2010) · Communications of the ACM · doi:10.1145/1721654.1721679 · Cited by 229
Generalization and Parameter Estimation in Feedforward Nets: Some Experiments
Morgan, N., & Bourlard, H. (1989) · neural information processing systems · Cited by 225
Ren, B., Ge, S. S., Su, C.-Y., & Lee, T. H. (2008) · IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics) · doi:10.1109/tsmcb.2008.2006368 · Cited by 214
On the safety of machine learning: Cyber-physical systems, decision sciences, and data products
Varshney, K. R., & Alemzadeh, H. (2017) · Big Data · doi:10.1089/big.2016.0051 · Cited by 211
Tani, J., Ito, M., & Sugita, Y. (2004) · Neural Networks · doi:10.1016/j.neunet.2004.05.007 · Cited by 208
Towards connected autonomous driving: review of use-cases
Montanaro, U., Dixit, S., Fallah, S., Dianati, M., Stevens, A., Oxtoby, D., & Mouzakitis, A. (2018) · Vehicle System Dynamics · doi:10.1080/00423114.2018.1492142 · Cited by 208 · cited 1×
“Autonomous vehicles have gained significant interest as solutions to these challenges [4]–[7].”
BackgroundSelf-citationTrajectory Planning for Autonomous High-Speed Overtaking in Structured Environments Using Robust MPC
Dixit, S., Montanaro, U., Dianati, M., Oxtoby, D., Mizutani, T., Mouzakitis, A., & Fallah, S. (2019) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2019.2916354 · Cited by 206 · cited 1×
“Other avenues for future research include learning driving manoeuvrers which are still typically achieved through classical control techniques, such as overtaking [227], [228] or merging [229], [230].”
Self-citationDriving Policy Transfer via Modularity and AbstractionPreprint
Müller, M., Dosovitskiy, A., Ghanem, B., & Koltun, V. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1804.09364 · Cited by 200 · cited 1×
“ferring driving policies from simulation to the real world easier [143], [144].”
BackgroundHybrid Reward Architecture for Reinforcement LearningPreprint
van Seijen, H., Fatemi, M., Romoff, J., Laroche, R., Barnes, T., & Tsang, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1706.04208 · Cited by 196 · cited 1×
“One solution to this is a Hybrid Reward Architecture [177], where the system uses a decomposed reward function and learns a separate value function for each component reward function.”
MethodEnd-to-End Learning of Driving Models with Surround-View Cameras and Route Planners
Hecker, S., Dai, D., & Van Gool, L. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01234-2_27 · Cited by 191 · cited 2×
“[137] Supervised Learning CNN 360-degree view camera image, route plan Steering angle, velocity Takes route plan into account Lack of live testing No live testing, tested on data set image examples only”
MethodEnd-to-End Deep Reinforcement Learning for Lane Keeping AssistPreprint
Sallab, A. E., Abdou, M., Perot, E., & Yogamani, S. K. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.04340 · Cited by 187 · cited 2×
“The lane keeping performance of (a) the DQN with discretised outputs and (b) DDAC with continuous output values [122].”
BackgroundLearning Visual Odometry with a Convolutional Network
Konda, K., & Memisevic, R. (2015) · International Conference on Computer Vision Theory and Applications · doi:10.5220/0005299304860490 · Cited by 170
Large-scale cost function learning for path planning using deep inverse reinforcement learning
Wulfmeier, M., Rao, D., Wang, D. Z., Ondrúška, P., & Posner, I. (2017) · The International Journal of Robotics Research · doi:10.1177/0278364917722396 · Cited by 170 · cited 1×
“[133] proposed an IRL approach for deep learning.”
BackgroundRobot Learning in Homes: Improving Generalization and Reducing Dataset BiasPreprint
Gupta, A., Murali, A., Gandhi, D., & Pinto, L. (2018) · Neural Information Processing Systems · doi:10.48550/arxiv.1807.07049 · Cited by 170 · cited 1×
“For complex tasks, such as autonomous driving, the diversity of the data set should be ensured if the aim is to train a generalisable model which can drive in all different environments [56], [57].”
BackgroundContinuous Inverse Optimal Control with Locally Optimal ExamplesPreprint
Levine, S., & Koltun, V. (2012) · arXiv (Cornell University) · doi:10.48550/arxiv.1206.4617 · Cited by 165 · cited 1×
“There are some solutions to minimising the effect of suboptimal demonstrations; using multiple trajectories and averaging over multiple sets to find a reward function or removing the assumption of global optimality [135].”
BackgroundAutonomous braking system via deep reinforcement learning
Chae, H., Kang, C. M., Kim, B., Kim, J., Chung, C. C., & Choi, J. W. (2017) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.2017.8317839 · Cited by 162 · cited 3×
“[111], who proposed an autonomous braking”
BackgroundMethodInfluentialCausal Confusion in Imitation LearningPreprint
de Haan, P., Jayaraman, D., & Levine, S. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.11979 · Cited by 161 · cited 1×
“assumption made by most learning algorithms [53]–[55].”
Transfer from Simulation to Real World through Learning Deep Inverse Dynamics ModelPreprint
Christiano, P. F., Shah, Z., Mordatch, I., Schneider, J., Blackwell, T., Tobin, J., Abbeel, P., & Zaremba, W. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1610.03518 · Cited by 159 · cited 1×
“policies from simulation to the real world [199]–[202].”
BackgroundSelf-Driving Cars and the Urban Challenge
Urmson, C., & Whittaker, W. (2008) · IEEE Intelligent Systems · doi:10.1109/mis.2008.34 · Cited by 157
Safe Exploration in Markov Decision Processes
Moldovan, T. M., & Abbeel, P. (2012) · arXiv · doi:10.48550/arxiv.1205.4810 · Cited by 152 · cited 1×
“Exploration poses safety challenges as the agent is encouraged to take random actions, which can lead to catastrophic events if not considered beforehand [210]–[213].”
BackgroundVehicle following control design for automated highway systems
Raza, H., & Ioannou, P. (1996) · IEEE Control Systems · doi:10.1109/mcs.1996.546270 · Cited by 150
Autonomous vehicle control systems — a review of decision making
Veres, S. M., Molnar, L., Lincoln, N. K., & Morice, C. P. (2011) · Proceedings of the Institution of Mechanical Engineers Part I Journal of Systems and Control Engineering · doi:10.1177/2041304110394727 · Cited by 143
Learning from Demonstration for Autonomous Navigation in Complex Unstructured Terrain
Silver, D., Bagnell, J. A., & Stentz, A. (2010) · The International Journal of Robotics Research · doi:10.1177/0278364910369715 · Cited by 142
Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Art
Janai, J., Güney, F., Behl, A., & Geiger, A. (2017) · arXiv · doi:10.48550/arxiv.1704.05519 · Cited by 141
Learning a deep neural net policy for end-to-end control of autonomous vehicles
Rausch, V., Hansen, A., Solowjow, E., Liu, C., Kreuzer, E. J., & Hedrick, J. K. (2017) · Proceedings of the ... American Control Conference/Proceedings of the American Control Conference · doi:10.23919/acc.2017.7963716 · Cited by 140
End-to-End Deep Learning for Steering Autonomous Vehicles Considering Temporal DependenciesPreprint
Eraqi, H. M., Moustafa, M. N., & Honer, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1710.03804 · Cited by 133 · cited 1×
“[93] utilised a Convolutional Long Short-Term Memory Recurrent Neural Network (C-LSTM) to learn to steer a vehicle based on visual and dynamic temporal”
MethodEnd-to-End Driving Via Conditional Imitation Learning
Codevilla, F., Muller, M., Lopez, A., Koltun, V., & Dosovitskiy, A. (2018) · 2018 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2018.8460487 · Cited by 129
Hierarchical Apprenticeship Learning with Application to Quadruped Locomotion
Kolter, J. Z., Abbeel, P., & Ng, A. Y. (2007) · Cited by 122
An Approach to Tune Fuzzy Controllers Based on Reinforcement Learning for Autonomous Vehicle Control
Dai, X., Li, C. K., & Rad, A. B. (2005) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2005.853698 · Cited by 121 · cited 2×
“[107], this reward function does not consider passenger comfort which could lead to harsh accelerations or decelerations.”
BackgroundMethodQuery-Efficient Imitation Learning for End-to-End Autonomous Driving
Zhang, J., & Cho, K. (2016) · arXiv · doi:10.48550/arxiv.1605.06450 · Cited by 118 · cited 1×
“[125] proposed a supervised learning method, SafeDAgger, for training a CNN to drive in a TORCS simulation.”
Combining neural networks and tree search for task and motion planning in challenging environments
Paxton, C., Raman, V., Hager, G. D., & Kobilarov, M. (2017) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8206505 · Cited by 116 · cited 1×
“[141] where the high-level command is provided by another DNN responsible for decision making.”
BackgroundParameterized Batch Reinforcement Learning for Longitudinal Control of Autonomous Land Vehicles
Huang, Z., Xu, X., He, H., Tan, J. H., & Sun, Z. (2017) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2017.2712561 · Cited by 107 · cited 1×
“[108] presented a Parameterised Batch Actor-Critic (PBAC) reinforcement learning algorithm for longitudinal control of autonomous vehicles based on actor-critic algorithms.”
MethodOn Offline Evaluation of Vision-Based Driving Models
Codevilla, F., López, A. M., Koltun, V., & Dosovitskiy, A. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01267-0_15 · Cited by 103
ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the WorstPreprint
Bansal, M., Krizhevsky, A., & Ogale, A. S. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.03079 · Cited by 102 · cited 2×
“Moving away from end-to-end approaches, researchers at Waymo recently presented ChauffeurNet [142].”
BackgroundDeep Object-Centric Policies for Autonomous Driving
Wang, D., Devin, C., Cai, Q.-Z., Yu, F., & Darrell, T. (2019) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2019.8794224 · Cited by 102 · cited 1×
“[128] demonstrated that DAgger can be used to train an object-centric policy, which uses salient objects in the image (e.”
BackgroundNeural Network Vision for Robot Driving
Pomerleau, D. (1997) · Springer eBooks · doi:10.1007/978-1-4615-6325-9_4 · Cited by 101 · cited 1×
“One of the earliest applications of artificial neural networks to the vehicle control problem was the Autonomous Land Vehicle in a Neural Network (ALVINN) system by Pomerleau in 1989 which was first described in [83] and further extended in [84].”
BackgroundImitation learning for locomotion and manipulation
Ratliff, N., Bagnell, J. A., & Srinivasa, S. S. (2007) · IEEE-RAS International Conference on Humanoid Robots (Humanoids) · doi:10.1109/ichr.2007.4813899 · Cited by 100 · cited 1×
“function from expert demonstrations of the task [180]–[183].”
BackgroundAn Analysis of ISO 26262: Using Machine Learning Safely in Automotive SoftwarePreprint
Salay, R., Queiroz, R., & Czarnecki, K. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1709.02435 · Cited by 99
Deep neural networks are robust to weight binarization and other non-linear distortionsPreprint
Merolla, P. A., Appuswamy, R., Arthur, J. V., Esser, S. K., & Modha, D. S. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.01981 · Cited by 98 · cited 1×
“Similarly, imposing constraints on the network weights through weight clipping has also been shown to improve robustness [188], [189].”
BackgroundLearning from Demonstrations for Real World Reinforcement LearningPreprint
Hester, T., Vecerík, M., Pietquin, O., Lanctot, M., Schaul, T., Piot, B., Sendonaris, A., Dulac-Arnold, G., Osband, I., Agapiou, J., et al. (2017) · arXiv (Cornell University) · Cited by 96
Deep Learning in Automotive Software
Falcini, F., Lami, G., & Costanza, A. M. (2017) · IEEE Software · doi:10.1109/ms.2017.79 · Cited by 96 · cited 1×
“This reveals further need for these standards to be revised to address machine learning systems for autonomous vehicles [226].”
BackgroundFull-range adaptive cruise control based on supervised adaptive dynamic programming
Zhao, D., Hu, Z., Xia, Z., Alippi, C., Zhu, Y., & Wang, D. (2013) · Neurocomputing · doi:10.1016/j.neucom.2012.09.034 · Cited by 93
A Control Strategy of Autonomous Vehicles Based on Deep Reinforcement Learning
Xia, W., Li, H., & Li, B. (2016) · International Symposium on Computational Intelligence and Design · doi:10.1109/iscid.2016.2054 · Cited by 93 · cited 2×
“[120] Supervised Reinforcement Learning Feedforward network with 2 hidden layers Not mentioned Steering, acceleration, braking Fast training Unstable (Can steer off the road) Simulation”
BackgroundMethodTheory and application of reward shaping in reinforcement learning
DEJONG, G. F., & Laud, A. (2004) · Cited by 92
Yin, H., & Berger, C. (2017) · 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.2017.8317828 · Cited by 79 · cited 2×
“For a more complete overview of available autonomous driving data sets, see the survey by Yin & Berger [78].”
BackgroundUsing Local Trajectory Optimizers to Speed Up Global Optimization in Dynamic Programming
Atkeson, C. G. (1993) · Cited by 78
A supervised Actor–Critic approach for adaptive cruise control
Zhao, D., Wang, B., & Liu, D. (2013) · Soft Computing · doi:10.1007/s00500-013-1110-y · Cited by 77 · cited 2×
“Also, the highly non-linear nature of driving means that control methods based on linearisation of the vehicle model or other algebraic analytical solutions are often infeasible or do not scale well [22], [23].”
BackgroundSelf-Learning Cruise Control Using Kernel-Based Least Squares Policy Iteration
Wang, J., Xu, X., Liu, D., Sun, Z., & Chen, Q. (2013) · IEEE Transactions on Control Systems Technology · doi:10.1109/tcst.2013.2271276 · Cited by 76 · cited 1×
“trollers trained with reinforcement learning [109], [110].”
Combining Deep Reinforcement Learning and Safety Based Control for Autonomous DrivingPreprint
Xiong, X., Wang, J., Zhang, F., & Li, K. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.00147 · Cited by 76 · cited 1×
“[214] suggested a control structure which combines reinforcement learning based control with safety based control and path tracking.”
BackgroundExploiting Model Uncertainty Estimates for Safe Dynamic Control Learning
Schneider, J. (1996) · Cited by 75
Evaluation of Policy Gradient Methods and Variants on the Cart-Pole Benchmark
Riedmiller, M., Peters, J., & Schaal, S. (2007) · doi:10.1109/adprl.2007.368196 · Cited by 75
The 1997 automated highway free agent demonstration
Thorpe, C., Jochem, T. M., & Pomerleau, D. (2002) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.1997.660524 · Cited by 69
LIDAR-based driving path generation using fully convolutional neural networks
Caltagirone, L., Bellone, M., Svensson, L., & Wahde, M. (2017) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.2017.8317618 · Cited by 68
Fuzzylot: a novel self-organising fuzzy-neural rule-based pilot system for automated vehicles
Pasquier, M., Quek, C. H., & Toh, M. (2001) · Neural Networks · doi:10.1016/s0893-6080(01)00048-x · Cited by 64 · cited 1×
“Control of these autonomous vehicles was handled via rule-based controllers, where the parameters are set by the developers and hand-tuned after simulation and field testing [17]–[19].”
MethodTowards Adapting Deep Visuomotor Representations from Simulated to Real Environments.Preprint
Tzeng, E., Devin, C., Hoffman, J., Finn, C., Peng, X., Levine, S., Saenko, K., & Darrell, T. (2015) · arXiv (Cornell University) · Cited by 61
A review of inverse reinforcement learning theory and recent advances
Shao, Z., & Joo, E. M. (2012) · IEEE Congress on Evolutionary Computation (CEC) · doi:10.1109/cec.2012.6256507 · Cited by 57
Trajectory Planning for Autonomous High-Speed Overtaking using MPC with Terminal Set Constraints
Dixit, S., Montanaro, U., Fallah, S., Dianati, M., Oxtoby, D., Mizutani, T., & Mouzakitis, A. (2018) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.2018.8569529 · Cited by 55
Self-citationYang, S., Wang, W., Liu, C., Deng, W., & Hedrick, J. K. (2017) · 2017 IEEE Intelligent Vehicles Symposium (IV) · doi:10.1109/ivs.2017.7995850 · Cited by 54 · cited 1×
“removing unnecessary data from the training and system input data [157].”
BackgroundLearning Autonomous Driving Styles and Maneuvers from Expert Demonstration
Silver, D., Bagnell, J. A., & Stentz, A. (2013) · Springer tracts in advanced robotics · doi:10.1007/978-3-319-00065-7_26 · Cited by 53 · cited 2×
“such rule-based controllers to generalise to new scenarios [21].”
BackgroundMethodA Study on Run Time Assurance for Complex Cyber Physical Systems
Clark, M., Koutsoukos, X., Porter, J., Kumar, R., Pappas, G. J., Sokolsky, O. V., Lee, I., & Pike, L. (2013) · Defense Technical Information Center (DTIC) · doi:10.21236/ada585474 · Cited by 52
Ross, P. E. (2014) · IEEE Spectrum · doi:10.1109/mspec.2014.6821623 · Cited by 51 · cited 1×
“autonomous vehicles in terms of better fuel economy [9], [10], reduced pollution, car sharing [11], increased productivity, and”
BackgroundJacklin, S. A., Schumann, J. M. P., Gupta, P., Richard, M., Guenther, K., & Soares, F. (2005) · AIAA Infotech@Aerospace Conference · doi:10.2514/6.2005-6912 · Cited by 51
Toward autonomous driving: the CMU Navlab. II. Architecture and systems
Thorpe, C., Herbert, M., Takeo Kanade, A., & Shafter, S. (1991) · IEEE Expert · doi:10.1109/64.85920 · Cited by 49
Zhao, D., Xia, Z., & Zhang, Q. (2017) · IEEE Computational Intelligence Magazine · doi:10.1109/mci.2017.2670463 · Cited by 49 · cited 1×
“(b) Velocity profiles of the lead vehicle and the three controllers [113].”
BackgroundA learning model for personalized adaptive cruise control
Chen, X., Zhai, Y., Lu, C., Gong, J., & Wang, G. (2017) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2017.7995748 · Cited by 36
Challenges in the Verification of Reinforcement Learning Algorithms
van Wesel, P., & Goodloe, A. E. (2017) · NASA STI Repository (National Aeronautics and Space Administration) · Cited by 30
Research works printed but not linked: 26
- ripley, 1996, Artificial Neural Networks
- schalkoff, 1997, Artificial Neural Networks
- ng, 2004, Feature selection, L1 vs. L2 regularization, and rotational invariance, Proc 21st Int Conf Mach Learn (ICML), 78
- 10.1126/science.153.3731.34 · DOI · Paper page
- barron, 2009, Parallelizing Reinforcement Learning
- abraham, 2017, Consumer Interest in Automation Preliminary Observations Exploring a Year’s Change
- knight, 2016, An Ambitious Plan to Build a Self-Driving Borg
- 2016, Cooperative Intelligent Transportation Systems—Research Theme Analysis Report
- 2017, Consumer Acceptance of Autonomous Vehicles
- bosankic, 2017, How consumers’ Perception of Autonomous Cars Will Influence Their Adoption
- sevchuk, 2016, Hyperparameter Optimization for Neural Networks
- raffel, 2015, Neural Network Hyperparameters
- 2018, Global status report on road safety 2018
- 2019, Waymo open dataset An autonomous driving dataset
- 2018, Udacity Self-Driving Car Dataset
- 2018, Autonomous car development platform from nvidia drive px2
- williams, 1987, Reinforcement learning connectionist systems
- 2018, Cyclone V—Overview
- 2018, Evolution of the Eye
- 0, CarSim
- nguyen, 2015, Deep neural networks are easily fooled, Proc IEEE Conf Comput Vis Pattern Recognit (CVPR), 427
- 2017, 2016 road safety statistics What is behind the figures
- 2017, 2016 Fatal Motor Vehicle Crashes Overview
- maurer, 2016, Autonomous Driving
- 2017, Research on the Impacts of Connected and Autonomous Vehicles (CAVs) on Traffic Flow Summary Report
- 2019, The Open Racing Car Simulator
Cited by
Sorted by citations
Review of deep learning: concepts, CNN architectures, challenges, applications, future directions
Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A., Duan, Y., Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., & Farhan, L. (2021) · Journal of Big Data · doi:10.1186/s40537-021-00444-8 · Cited by 7,906
Ahmad, K., Maabreh, M., Ghaly, M., Khan, K., Qadir, J., & Al-Fuqaha, A. (2022) · Computer Science Review · doi:10.1016/j.cosrev.2021.100452 · Cited by 291
“…autonomous cars where one of the key responsibilities of the AI module is continuous monitoring of the surrounding environment and the prediction of di ff erent events, which generally involves the detection and recognition of various objects such as pedestrians, vehicles, and roadside objects [38].”
BackgroundA Survey of Deep RL and IL for Autonomous Driving Policy Learning
Zhu, Z., & Zhao, H. (2021) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2021.3134702 · Cited by 260
“References [4] addresses the deep learning techniques for AD with a focus on perception and control, while [32] addresses control only.”
BackgroundMachine learning for next-generation intelligent transportation systems: A survey
Yuan, T., Da Rocha Neto, W., Rothenberg, C. E., Obraczka, K., Barakat, C., & Turletti, T. (2022) · Transactions on Emerging Telecommunications Technologies · doi:10.1002/ett.4427 · Cited by 179
“…works include [6-16], which focus on specific aspects of ITS, such as vehicular networks [7], vehicle detection [8-10], safety of vehicular ad-hoc network (VANET) [6, 15], VANET performance optimization [16], autonomous vehicle control [13], and Deep Reinforcement Learning (DRL) for 6G VANETs [14].”
Elallid, B. B., Benamar, N., Hafid, A. S., Rachidi, T., & Mrani, N. (2022) · Journal of King Saud University - Computer and Information Sciences · doi:10.1016/j.jksuci.2022.03.013 · Cited by 179
“Kuutti et al. (2020) discussed DL methods for controlling AVs and its promising performance in complex scenarios.”
Metadata
- Authors
- Kuutti, S., Bowden, R., Jin, Y., Barber, P., & Fallah, S.
- Affiliations
- University of Surrey (GB)
- Jaguar Land Rover (United Kingdom) (GB)
- Venue
- IEEE Transactions on Intelligent Transportation Systems
- Published
- 2021-02
- Volume
- 22
- Issue
- 2
- Pages
- 712-733
- ISSN
- 1524-9050 (ISSN-L)
- 1558-0016
- OpenAlex
- W2995697759
- Cited by
- 96
- References
- 230
- Citations by year
- 2026: 12 · 2025: 7 · 2024: 16 · 2023: 21 · 2022: 14 · 2021: 21
- FWCI
- 5.64
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Intelligent Transportation Systems (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Surrey Research Insight Open Access (The University of Surrey) (submittedVersion)
- PUB – Publications at Bielefeld University (Bielefeld University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- U.K.-Engineering and Physical Sciences Research Council (EP/R512217/1)
- Jaguar Land Rover
- Topics
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Physical Sciences › Engineering › Automotive Engineering › Vehicle Dynamics and Control Systems
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Keywords
- autonomous vehicles, vehicle control, deep learning control, verification and validation, safety, goal specification, intelligent transportation systems
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
