Filters








8,360 Hits in 7.2 sec

Imitating Driver Behavior with Generative Adversarial Networks [article]

Alex Kuefler, Jeremy Morton, Tim Wheeler, Mykel Kochenderfer
2017 arXiv   pre-print
The ability to accurately predict and simulate human driving behavior is critical for the development of intelligent transportation systems.  ...  We extend Generative Adversarial Imitation Learning to the training of recurrent policies, and we demonstrate that our model outperforms rule-based controllers and maximum likelihood models in realistic  ...  Whereas Generative Adversarial Imitation Learning captures human-like behavior present in the dataset, simulators may also wish to enforce certain behaviors (e.g., explicitly modeling driver style) by  ... 
arXiv:1701.06699v1 fatcat:elhdfvqu75bjjbtqav4eqhwqse

Modeling Adversarial Behavior Against Mobility Data Privacy

Roberto Pellungrini, Luca Pappalardo, Filippo Simini, Anna Monreale
2020 IEEE transactions on intelligent transportation systems (Print)  
We use simulated annealing to optimize the movement of the adversary and simulate a possible attack on mobility data.  ...  We finally test the effectiveness of our approach on real human mobility data, showing that it can simulate the knowledge gathering process for an adversary in a more realistic way.  ...  In this article, we propose a data-driven approach to realistically simulate the behavior of a malicious adversary in the acquisition of background knowledge for privacy attacks in mobility data.  ... 
doi:10.1109/tits.2020.3021911 fatcat:i7krbpufcnhudcunzukwz5yph4

Generative Adversarial Networks Synthesize Realistic OCT Images of the Retina [article]

Stephen G. Odaibo, M.D., M.S.
2019 arXiv   pre-print
We report, to our knowledge, the first end-to-end application of Generative Adversarial Networks (GANs) towards the synthesis of Optical Coherence Tomography (OCT) images of the retina.  ...  Generative models have gained recent attention for the increasingly realistic images they can synthesize, given a sampling of a data type.  ...  Chang of IBM (retired) for endorsement to the arXiv's Computer Science Artificial Intelligence Section. And he thanks  ... 
arXiv:1902.06676v1 fatcat:vptm47pzcbhyrpxfy7metei4pi

Photorealism in Driving Simulations: Blending Generative Adversarial Image Synthesis with Rendering [article]

Ekim Yurtsever, Dongfang Yang, Ibrahim Mert Koc, Keith A. Redmill
2022 arXiv   pre-print
Driving simulators play a large role in developing and testing new intelligent vehicle systems.  ...  Given a 3D scene, we partially render only important objects of interest, such as vehicles, and use generative adversarial processes to synthesize the background and the rest of the image.  ...  PRELIMINARIES Generative Adversarial Networks (GAN) [9] use a generator G and a discriminator D in a simultaneous adversarial training strategy.  ... 
arXiv:2007.15820v2 fatcat:pkpckde4jvepfmw7yifvrsblom

Network Environment Design for Autonomous Cyberdefense [article]

Andres Molina-Markham, Cory Miniter, Becky Powell, Ahmad Ridley
2021 arXiv   pre-print
To develop such understanding, it is necessary to develop RL agents using simulation and emulation systems allowing researchers to model a broad class of realistic threats and network conditions.  ...  Our framework enables the development and simulation of adversaries with sophisticated behavior that includes poisoning and evasion attacks on RL network defenders.  ...  Simulating network-defender experiences requires a large number of general-purpose computations.  ... 
arXiv:2103.07583v1 fatcat:entwvffalvgphhpgazjtum4djq

A Gaming Environment for Resilient Network Design and Adversarial Co-Evolution Modeling

YMarco Carvalho, Adrian Granados, James McLane, Evan Stoner
2014 Lecture Notes on Information Theory  
In this work, we propose that serious games can be used to engage humans in high-fidelity adversarial simulations that will allow us to capture, and possibly model the co-evolutionary behavior of adversaries  ...  While generally useful to provide an initial baseline, such static approaches fail to take into account the adaptive nature of the attacker.  ...  In this work, we propose that serious games can be used to engage human in high-fidelity adversarial simulations that will allow us to capture, and possibly model the co-evolution of behavior, and will  ... 
doi:10.12720/lnit.2.1.92-97 fatcat:2jcaq743hjc6lf6a3qykbj3oae

Challenges and Characteristics of Intelligent Autonomy for Internet of Battle Things in Highly Adversarial Environments [article]

Alexander Kott
2018 arXiv   pre-print
This paper explores the characteristics, capabilities and intelligence required of such a network of intelligent things and humans - Internet of Battle Things (IOBT).  ...  Numerous, artificially intelligent, networked things will populate the battlefield of the future, operating in close collaboration with human warfighters, and fighting as teams in highly adversarial environments  ...  A possible approach to developing the necessary capabilitiesboth human and AIis to train a human-agent team in immersive artificial environments.  ... 
arXiv:1803.11256v2 fatcat:ylkm3f3idnglnkefrw3frsksgu

CALEB: A Conditional Adversarial Learning Framework to Enhance Bot Detection [article]

George Dialektakis, Ilias Dimitriadis, Athena Vakali
2022 arXiv   pre-print
on the Conditional Generative Adversarial Network (CGAN) and its extension, Auxiliary Classifier GAN (AC-GAN), to simulate bot evolution by creating realistic synthetic instances of different bot types  ...  As highlighted by other researchers, most of these bots have malicious purposes and tend to mimic human behavior, posing high-level security threats on OSN platforms.  ...  Synthetic Bot Data Generation using GANs Generative Adversarial Networks (GANs) have recently been introduced as a data augmentation technique, where a GAN is trained to generate realistic synthetic samples  ... 
arXiv:2205.15707v1 fatcat:fvd26crb6jgstkjyrrk3q22k2a

A Survey on Safety-Critical Driving Scenario Generation – A Methodological Perspective [article]

Wenhao Ding, Chejian Xu, Mansur Arief, Haohong Lin, Bo Li, Ding Zhao
2022 arXiv   pre-print
Then, we discuss useful tools for scenario generation, including simulation platforms and packages.  ...  Autonomous driving systems have witnessed a significant development during the past years thanks to the advance in machine learning-enabled sensing and decision-making algorithms.  ...  A 2D platform named SMARTS is developed in [72] containing multiple diverse behavior models using both rule-based and learning-based models. [73] proposes MetaDrive, a 3D simulator that supports different  ... 
arXiv:2202.02215v4 fatcat:uxcvqxk6qna5jh53dwdwqanc4q

A next-generation platform for Cyber Range-as-a-Service [article]

Vittorio Orbinato
2021 arXiv   pre-print
In the last years, Cyber Ranges have become a widespread solution to train professionals for responding to cyber threats and attacks.  ...  automatic monitoring of the participants' activities, and the emulation of their behavior.  ...  Adversary emulation is useful to measure the ability of the network itself to identify threats and prevent damages to its devices.  ... 
arXiv:2112.11233v1 fatcat:hercjvwlkzgybay737edulu6ja

Intelligent Autonomous Things on the Battlefield [article]

Alexander Kott, Ethan Stump
2019 arXiv   pre-print
This chapter explores the characteristics, capabilities and intelli-gence required of such a network of intelligent things and humans - Internet of Battle Things (IOBT).  ...  Numerous, artificially intelligent, networked things will populate the battlefield of the future, operating in close collaboration with human warfighters, and fighting as teams in highly adversarial environments  ...  Humans have developed many control schemes that allow us to bend natural processes to our will, and machines must continue these developments, tapping into a capacity for learning and adaptation to magnify  ... 
arXiv:1902.10086v1 fatcat:uxdjhtr2jnemfcmuoputig273y

Systematic review of features for co‐simulating security incidents in Cyber‐Physical Systems

Ricardo M. Czekster, Charles Morisset, John A. Clark, Sadegh Soudjani, Charalampos Patsios, Peter Davison
2021 Security and Privacy  
Addressing security requires developing modeling and simulation tools that approximate and replicate adversarial behavior in the SG.  ...  Cyber-physical security is a major source of concern due to the high reliance of the SG on Information and Communication Technologies (ICT) and their widespread use.  ...  ACKNOWLEDGMENTS The authors wish to acknowledge funding from the Industrial Strategy Challenge Fund and Engineering and Physical Sciences Research Council (EPSRC/UK) EP/V012053/1 for The Active Building  ... 
doi:10.1002/spy2.150 fatcat:mcrafuc2tnbqnmsywuniygcbcq

Modeling multiple communities of interest for interactive simulation and gaming: the dynamic adversarial gaming algorithm project

Eugene Santos, Jr., Qunhua Zhao, Felicia Pratto, Adam R. Pearson, Bruce McQueary, Andy Breeden, Lee Krause, Kevin Schum, Dawn A. Trevisani
2007 Modeling and Simulation for Military Operations II  
The Dynamic Adversarial Gaming Algorithm (DAGA) project aims to provide a wargaming environment for automation of simulating dynamics of geopolitical crisis and eventually be applied to military simulation  ...  We do so by representing cultural variables and their influence on behavior using probabilistic networks.  ...  system to produce simulation environment that is more realistic.  ... 
doi:10.1117/12.722319 fatcat:avz6sxt5q5eerlxqc2stxy36aq

AdvSim: Generating Safety-Critical Scenarios for Self-Driving Vehicles [article]

Jingkang Wang, Ava Pun, James Tu, Sivabalan Manivasagam, Abbas Sadat, Sergio Casas, Mengye Ren, Raquel Urtasun
2022 arXiv   pre-print
In this paper, we propose AdvSim, an adversarial framework to generate safety-critical scenarios for any LiDAR-based autonomy system.  ...  Traditionally, those scenarios are generated for a few scenes with respect to the planning module that takes ground-truth actor states as input.  ...  We then describe how we parameterize the adversarial actors' behaviors (Sec 3.2) and conduct realistic LiDAR simulation to generate new LiDAR sweeps (Sec 3.3).  ... 
arXiv:2101.06549v3 fatcat:xuhsg67ybbev3ma6ux352wxcea

Towards digital cognitive clones for the decision-makers: adversarial training experiments

Mariia Golovianko, Svitlana Gryshko, Vagan Terziyan, Tuure Tuunanen
2021 Procedia Computer Science  
The digital component of the environment contains special modifications of Generative Adversarial Networks, which include a human-operator as a trainer, an autonomous Pi-Mind clone as a trainee (a discriminator  ...  The digital component of the environment contains special modifications of Generative Adversarial Networks, which include a human-operator as a trainer, an autonomous Pi-Mind clone as a trainee (a discriminator  ...   Customer experience, supply chain and asset management: simulations with the self-managed "digital customers" and digitalized processes can be used to obtain the optimal settings in real-world environments  ... 
doi:10.1016/j.procs.2021.01.155 fatcat:74qe4vvcurbftco53z45baupoq
« Previous Showing results 1 — 15 out of 8,360 results