Filters








1,930 Hits in 8.0 sec

Imagining The Road Ahead: Multi-Agent Trajectory Prediction via Differentiable Simulation [article]

Adam Scibior, Vasileios Lioutas, Daniele Reda, Peyman Bateni, Frank Wood
2021 arXiv   pre-print
We develop a deep generative model built on a fully differentiable simulator for multi-agent trajectory prediction.  ...  We name our model ITRA, for "Imagining the Road Ahead".  ...  We acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canada CIFAR AI Chairs Program, and the Intel Parallel Computing Centers program.  ... 
arXiv:2104.11212v1 fatcat:eqmbvfttyfbejclwti7d2ct3li

Do Androids Dream of Electric Fences? Safety-Aware Reinforcement Learning with Latent Shielding [article]

Peter He, Borja G. Leon, Francesco Belardinelli
2021 arXiv   pre-print
Latent shielding leverages internal representations of the environment learnt by model-based agents to "imagine" future trajectories and avoid those deemed unsafe.  ...  In recent years, a variety of approaches have been put forward to address the challenges of safety-aware reinforcement learning; however, these methods often either require a handcrafted model of the environment  ...  Imagination-Augmented Agents for Deep Re- Topcu, U.; and Feng, L. 2021. Safe Multi-Agent Reinforce- inforcement Learning. In Proceedings of the 31st Interna- ment Learning via Shielding.  ... 
arXiv:2112.11490v1 fatcat:o5curea7ena5blafok3uldvfni

A Review of Tracking, Prediction and Decision Making Methods for Autonomous Driving [article]

Florin Leon, Marius Gavrilescu
2019 arXiv   pre-print
), prediction (predicting the future motion of surrounding vehicles in order to navigate through various traffic scenarios) and decision making (analyzing the available actions of the ego car and their  ...  consequences to the entire driving context).  ...  Acknowledgements We kindly thank Continental AG for their great cooperation within Proreta 5, which is a joint research project of the Technical University of Darmstadt, University of Bremen, Technical  ... 
arXiv:1909.07707v1 fatcat:h2ttehcuzrc2dnnmqzigilcyri

A Mental Simulation Approach for Learning Neural-Network Predictive Control (in Self-Driving Cars)

Mauro Da Lio, Riccardo Dona, Gastone Pietro Rosati Papini, Francesco Biral, Henrik Svensson
2020 IEEE Access  
lane 50 m ahead.  ...  Then, offline (at sleep/dream state), the same agent synthesizes "inverse models" for control and behaviors via, respectively, embodied simulations (low-level detailed simulations) and episodic simulations  ... 
doi:10.1109/access.2020.3032780 fatcat:4tzdpitfznampobz7fu5q34g4m

Airborne Collision Avoidance System as a Cyber-Physical System
English

NAE Andrei C., DUMITRACHE Ioan
2015 INCAS Bulletin  
The new system will represent a step change over the performance of current technology.  ...  Here we consider the Air Transportations System of the future as a Cyber-Physical System and analyze the implications of doing so from different perspectives.  ...  Any aircraft that has the capability to compute/predict its avoidance trajectory at least 5 seconds ahead and has the capability to communicate that information via a data link, can be protected with ACAS  ... 
doi:10.13111/2066-8201.2015.7.4.12 fatcat:3k47p237wzcadbwtlpez55eg5e

Learning Dynamics Model in Reinforcement Learning by Incorporating the Long Term Future [article]

Nan Rosemary Ke, Amanpreet Singh, Ahmed Touati, Anirudh Goyal, Yoshua Bengio, Devi Parikh, Dhruv Batra
2019 arXiv   pre-print
An exploration strategy can be devised by searching for unlikely trajectories under the model.  ...  In model-based reinforcement learning, the agent interleaves between model learning and planning. These two components are inextricably intertwined.  ...  We are also grateful to the reviewers for their constructive feedback which helped to improve the clarity of the paper.  ... 
arXiv:1903.01599v2 fatcat:qarnknfy5vcntl7v4b3suyw5ne

Mastering Atari with Discrete World Models [article]

Danijar Hafner, Timothy Lillicrap, Mohammad Norouzi, Jimmy Ba
2021 arXiv   pre-print
We introduce DreamerV2, a reinforcement learning agent that learns behaviors purely from predictions in the compact latent space of a powerful world model.  ...  With the same computational budget and wall-clock time, Dreamer V2 reaches 200M frames and surpasses the final performance of the top single-GPU agents IQN and Rainbow.  ...  Acknowledgements We thank our anonymous reviewers for their feedback and Nick Rhinehart for an insightful discussion about the potential benefits of categorical latent variables.  ... 
arXiv:2010.02193v3 fatcat:kpsruthmivcyxdab27sug4bgm4

A Review of Tracking and Trajectory Prediction Methods for Autonomous Driving

Florin Leon, Marius Gavrilescu
2021 Mathematics  
, i.e., anticipating the future trajectories and motion of other vehicles in order to facilitate navigating through various traffic conditions.  ...  Specifically, we focus on two aspects extensively explored in the related literature: tracking, i.e., identifying pedestrians, cars or obstacles from images, observations or sensor data, and prediction  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/math9060660 fatcat:qvikrr32tzd7fnjzs22u3ago4m

Continuum Traffic Simulation

J. Sewall, D. Wilkie, P. Merrell, M. C. Lin
2010 Computer graphics forum (Print)  
We present a novel method for the synthesis and animation of realistic traffic flows on large-scale road networks.  ...  , real-world road data.  ...  Acknowledgments The authors would like to thank Avneesh Sud at Microsoft and the members of the GAMMA group at UNC.  ... 
doi:10.1111/j.1467-8659.2009.01613.x fatcat:h4xds5l4fzcqfdq2txmltc44lq

Hierarchical Agent-Based Modeling for Improved Traffic Routing

Raghda Alqurashi, Tom Altman
2019 Applied Sciences  
The simulation results validate the ability of the proposed model and the included decision-making sub-models to both predict and improve the behaviors and intended actions of the agents.  ...  The model provides drivers with real-time alternative routes, computed via a decentralized multi-agent model, that tries to achieve a system-optimal traffic distribution within an entire system, thus reducing  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/app9204376 fatcat:f6ghgh3op5hbplcyitpo4jwsim

Leveraging human behavior models to predict paths in indoor environments

Bulent Tastan, Gita Sukthankar
2011 Pervasive and Mobile Computing  
One of the most powerful constraints governing many activity recognition problems is that imposed by the human actor.  ...  We demonstrate an approach for path prediction based on a model of visually guided steering that has been validated on human obstacle avoidance data.  ...  These script objects send their current locations via email. A multi-threaded Java email parser reads the email using the POP gmail interface and stores it in a database.  ... 
doi:10.1016/j.pmcj.2011.02.003 fatcat:ihaeiqm7rzgyvktrfivchzo3si

Learning a Decision Module by Imitating Driver's Control Behaviors [article]

Junning Huang, Sirui Xie, Jiankai Sun, Qiurui Ma, Chunxiao Liu, Jianping Shi, Dahua Lin, Bolei Zhou
2021 arXiv   pre-print
We show in the simulation experiments that our modular driving agent can generalize its driving decision and control to various complex scenarios where the rule-based programs fail.  ...  It can also generate smoother and safer driving trajectories than end-to-end neural policies.  ...  [25] imitate the mapping from predicted waypoints to control commands via neural network policy.  ... 
arXiv:1912.00191v3 fatcat:s34qyeugvzez7dpjbrwf3dklem

Closing the Planning-Learning Loop with Application to Autonomous Driving in a Crowd [article]

Panpan Cai, David Hsu
2021 arXiv   pre-print
Imagine an autonomous robot vehicle driving in dense, possibly unregulated urban traffic.  ...  We applied LeTSDrive to autonomous driving in crowded urban environments in simulation.  ...  [11] , multi-agent planning [12] , etc..  ... 
arXiv:2101.03834v2 fatcat:3rjws4a5h5grxf5rj7x6fxa34m

Traffic networks [chapter]

Kai Nagel
2004 Handbook of Graphs and Networks  
Another network aspect is the network of interaction between objects in the simulation, where these objects are not only travelers, but also traffic signals, traffic management centers, etc.  ...  Such models resemble typical molecular dynamics simulations, except that the spatial substrate is a graph instead of flat space, and particles are "intelligent".  ...  The Swiss Federal Administration provides the input data for the Switzerland studies.  ... 
doi:10.1002/3527602755.ch11 fatcat:5clkhcoiu5gzfkrcoidzdg6e54

Towards a Systematic Computational Framework for Modeling Multi-Agent Decision-Making at Micro Level for Smart Vehicles in a Smart World [article]

Qi Dai, Xunnong Xu, Wen Guo, Suzhou Huang, Dimitar Filev
2020 arXiv   pre-print
We propose a multi-agent based computational framework for modeling decision-making and strategic interaction at micro level for smart vehicles in a smart world.  ...  with finite look-ahead anticipation.  ...  Acknowledgement We thank Gint Puskorius and Jinhong Wang for several useful discussions and for their comments on the manuscript. We are also grateful to Paul Stieg for his help in literature review.  ... 
arXiv:2009.12213v1 fatcat:v26xpuhuqrgobcyccq4ql3z2lm
« Previous Showing results 1 — 15 out of 1,930 results