817 Hits in 4.5 sec

CARLA Real Traffic Scenarios – novel training ground and benchmark for autonomous driving [article]

Błażej Osiński, Piotr Miłoś, Adam Jakubowski, Paweł Zięcina, Michał Martyniak, Christopher Galias, Antonia Breuer, Silviu Homoceanu, Henryk Michalewski
2021 arXiv   pre-print
CRTS combines the realism of traffic scenarios and the flexibility of simulation. We use it to train agents using a reinforcement learning algorithm.  ...  This work introduces interactive traffic scenarios in the CARLA simulator, which are based on real-world traffic.  ...  We extensively used the Prometheus supercomputer, located in the Academic Computer Center Cyfronet in the AGH University of Science and Technology in Kraków, Poland.  ... 
arXiv:2012.11329v2 fatcat:euukpq5mprbupctlhvh3twbgpe

Cooperation-Aware Lane Change Maneuver in Dense Traffic based on Model Predictive Control with Recurrent Neural Network [article]

Sangjae Bae, Dhruv Saxena, Alireza Nakhaei, Chiho Choi, Kikuo Fujimura, Scott Moura
2019 arXiv   pre-print
This paper presents a real-time lane change control framework of autonomous driving in dense traffic, which exploits cooperative behaviors of other drivers.  ...  Finally, quantitative and qualitative analysis on simulation studies are presented to illustrate the benefits of the proposed framework.  ...  We utilize social GAN (SGAN) in [23] which efficiently and effectively captures multi-modal interactions between agents (drivers).  ... 
arXiv:1909.05665v2 fatcat:x4a4tgejnjftdmnowcovvmu77i

A Bayesian Driver Agent Model for Autonomous Vehicles System Based on Knowledge-Aware and Real-Time Data

Jichang Ma, Hui Xie, Kang Song, Hao Liu
2021 Sensors  
To explore the validity of the Bayesian driver agent model, we performed experiments on a driving simulation platform.  ...  The decision module based on dynamic Bayesian network (DBN) then makes an inferred decision using the traffic scene's feature values.  ...  In order to meet this objective, we used a multi-layer CNN to process the traffic scenario images, simulate the cognitive function region and predict the indicator values of the road scenes.  ... 
doi:10.3390/s21020331 pmid:33418987 fatcat:4wcdmmr32fcajfyl56calwb2vm

Potential Game Based Decision-Making Frameworks for Autonomous Driving [article]

Mushuang Liu, Ilya Kolmanovsky, H. Eric Tseng, Suzhou Huang, Dimitar Filev, Anouck Girard
2022 arXiv   pre-print
In addition, we provide cost function shaping approaches to constructing multi-agent potential games in autonomous driving.  ...  Decision-making for autonomous driving is challenging, considering the complex interactions among multiple traffic agents (e.g., autonomous vehicles (AVs), human drivers, and pedestrians) and the computational  ...  As each traffic agent has its own objective (also called self-interest), the multi-agent decisionmaking problem is intrinsically a multi-player game problem [7] , [12] - [15] .  ... 
arXiv:2201.06157v1 fatcat:hiqendwbyzg35osezmmjjnj7ey

Learning from All Vehicles [article]

Dian Chen, Philipp Krähenbühl
2022 arXiv   pre-print
We evaluate this system in closed-loop driving simulations.  ...  This system uses the behaviors of other agents to create more diverse driving scenarios without collecting additional data.  ...  Limitations and potential negative social impacts: Our approach is trained and evaluated in simulation alone and still incurs traffic infractions.  ... 
arXiv:2203.11934v2 fatcat:k3s65lcyyrfznpf5ojt3rrp6am

Learning to drive from a world on rails [article]

Dian Chen, Vladlen Koltun, Philipp Krähenbühl
2021 arXiv   pre-print
Our approach computes action-values for each training trajectory using a tabular dynamic-programming evaluation of the Bellman equations; these action-values in turn supervise the final vision-based driving  ...  To support learning from pre-recorded logs, we assume that the world is on rails, meaning neither the agent nor its actions influence the environment.  ...  The ego-vehicle stateL ego t = (x t , y t , v, θ) is always centered in this discretization. The position of the ego-vehicle (x t , y t ) is at the center of the spatial discretization.  ... 
arXiv:2105.00636v3 fatcat:j5luomy7sfchri6f5ksoseojo4

Agent Architecture for Adaptive Behaviours in Autonomous Driving

Mauro Da Lio, Riccardo Dona, Gastone Pietro Rosati Papini, Kevin Gurney
2020 IEEE Access  
This paper uses inspiration from these principles of organization in the design of an artificial agent for autonomous driving.  ...  The resulting Agent, developed in an EU H2020 Research and Innovation Action, is used to concretely demonstrate the emergence of adaptive behaviour with a significant level of autonomy.  ...  The ego car travels on a straight three-lane motorway following a vehicle that, unexpectedly drops an object (a traffic cone in the simulation, see schematic representation in Fig. 12 ).  ... 
doi:10.1109/access.2020.3007018 fatcat:fgqwypjyefcchh6fmn6hi4qjya

A Driver-Vehicle Model for ADS Scenario-based Testing [article]

Rodrigo Queiroz, Divit Sharma, Ricardo Caldas, Krzysztof Czarnecki, Sergio García, Thorsten Berger, Patrizio Pelliccione
2022 arXiv   pre-print
Our extensive evaluation shows the model's design effectiveness using NHTSA pre-crash scenarios, its motion realism in comparison to naturalistic urban traffic, and its scalability with traffic density  ...  The layered architecture of the model leverages behavior trees to express high-level behaviors in terms of lower-level maneuvers, affording multiple driving styles and reuse.  ...  *Scenario #16 required a map adaptation to perform correctly. mid-day traffic in Waterloo, Canada, which is part of the Waterloo Multi-Agent Traffic Dataset [48] .  ... 
arXiv:2205.02911v1 fatcat:m434pvjqhjhc3cg7vrvlph56s4

Deep Structured Reactive Planning [article]

Jerry Liu, Wenyuan Zeng, Raquel Urtasun, Ersin Yumer
2021 arXiv   pre-print
Through simulations based on both real-world driving and synthetically generated dense traffic, we demonstrate that our reactive model outperforms a non-reactive variant in successfully completing highly  ...  complex maneuvers (lane merges/turns in traffic) faster, without trading off collision rate.  ...  Our simulation settings involve both real-world traffic as well as synthetic dense traffic settings.  ... 
arXiv:2101.06832v2 fatcat:72xqeuv245hhlcgcye4ubenrvm

Survey of Deep Reinforcement Learning for Motion Planning of Autonomous Vehicles

Szilard Aradi
2020 IEEE transactions on intelligent transportation systems (Print)  
traffic.  ...  The paper describes vehicle models, simulation possibilities and computational requirements.  ...  Research Development and Innovation Office in the field of Artificial Intelligence (BME IE-MI-FM TKP2020).  ... 
doi:10.1109/tits.2020.3024655 fatcat:wk4c2ked3jho3jtqdn4o5ys4zu

End-to-End Deep Learning of Lane Detection and Path Prediction for Real-Time Autonomous Driving [article]

Der-Hau Lee, Jinn-Liang Liu
2021 arXiv   pre-print
in a host agent car driving along with other agents all in a real-time autonomous manner.  ...  DSUNet-PP outperforms UNet-PP in mean average errors of predicted curvature and lateral offset for path planning in dynamic simulation.  ...  affordance for direct perception in autonomous driving.  ... 
arXiv:2102.04738v2 fatcat:lso75tzjcrehhdtcd2wt6oxe6q

Conditional Affordance Learning for Driving in Urban Environments [article]

Axel Sauer, Nikolay Savinov, Andreas Geiger
2018 arXiv   pre-print
in simulation.  ...  In addition, our approach is the first to handle traffic lights and speed signs by using image-level labels only, as well as smooth car-following, resulting in a significant reduction of traffic accidents  ...  Perception We formulate perception as a multi-task learning (MTL) problem: using a single neural network we predict all affordances in a single forward pass.  ... 
arXiv:1806.06498v3 fatcat:vzxrg62hlve57lj3hvrjfgrgfy

End-to-End Multi-Task Deep Learning and Model Based Control Algorithm for Autonomous Driving [article]

Der-Hau Lee, Jinn-Liang Liu
2021 arXiv   pre-print
then identify the best one by two additional dynamic measures in real-time simulation.  ...  We also propose a learning- and model-based longitudinal controller using model predictive control method.  ...  The maximum speed of ego car (Ego) in [19] is 20 km/h in single-lane traffic. Cudrano et al.  ... 
arXiv:2112.08967v1 fatcat:xx4zmyuxp5gkpeijciyxjlpoai

Exploring the Limitations of Behavior Cloning for Autonomous Driving [article]

Felipe Codevilla, Eder Santana, Antonio M. López, Adrien Gaidon
2019 arXiv   pre-print
Behavior cloning in particular has been successfully used to learn simple visuomotor policies end-to-end, but scaling to the full spectrum of driving behaviors remains an unsolved problem.  ...  Driving requires reacting to a wide variety of complex environment conditions and agent behaviors. Explicitly modeling each possible scenario is unrealistic.  ...  ., traffic lights) and dynamic agents in the scene.  ... 
arXiv:1904.08980v1 fatcat:2736atclgfbhrhg72g2tzu3nwy

Virtual Scenario Simulation and Modeling Framework in Autonomous Driving Simulators

Mingyun Wen, Jisun Park, Yunsick Sung, Yong Woon Park, Kyungeun Cho
2021 Electronics  
In this study, a scenario simulation and modeling framework that simulates the behavior of objects that may be encountered while driving is proposed to address this problem.  ...  only be simulated up to a limited extent.  ...  In normal mode, normal scenarios in which human and animal agents are driven by motivation and all agents follow traffic rules are simulated.  ... 
doi:10.3390/electronics10060694 fatcat:vspefid4bvesfnfdjcvobstrqm
« Previous Showing results 1 — 15 out of 817 results