The Experts below are selected from a list of 146415 Experts worldwide ranked by ideXlab platform
Marius J Zollner - One of the best experts on this subject based on the ideXlab platform.
-
learning how to drive in a real World Simulation with deep q networks
IEEE Intelligent Vehicles Symposium, 2017Co-Authors: Peter Wolf, Christian Hubschneider, Michael Weber, Andre Bauer, Jonathan Hartl, Fabian Durr, Marius J ZollnerAbstract:We present a reinforcement learning approach using Deep Q-Networks to steer a vehicle in a 3D physics Simulation. Relying solely on camera image input the approach directly learns steering the vehicle in an end-to-end manner. The system is able to learn human driving behavior without the need of any labeled training data. An action-based reward function is proposed, which is motivated by a potential use in real World reinforcement learning scenarios. Compared to a naive distance-based reward function, it improves the overall driving behavior of the vehicle agent. The agent is even able to reach comparable to human driving performance on a previously unseen track in our Simulation environment.
David Sislak - One of the best experts on this subject based on the ideXlab platform.
-
a globe agent development platform with inaccessibility and mobility support
2005Co-Authors: David Sislak, Michal Pěchoucek, Martin Rehak, Milan Rollo, Dusan PavlicekAbstract:At present several Java-based multi-agent platforms from different developers are available, but none of them fully supports agent mobility and communication inaccessibility Simulation. They are thus unsuitable for experiments with large scale real-World Simulation. In this chapter we describe architecture of A-globe, fast, scalable and lightweight agent development platform with environmental Simulation and mobility support. Beside the functions common to most agent platforms it provides a position-based messaging service, so it can be used for experiments with extensive environment Simulation and communication inaccessibility. Simple benchmarks that compare the A-globe performance against other available agent platforms are also included.
-
a globe agent platform with inaccessibility and mobility support
Cooperative Information Agents, 2004Co-Authors: David Sislak, Milan Rollo, Michal PěchoucekAbstract:At present several Java-based multi-agent platforms from different developers are available, but none of them fully supports agent mobility and communication inaccessibility. They are thus no suitable for experiments with real-World Simulation. In this paper we describe architecture of newly developed agent platform \(\mathcal{A}\)-GLOBE. is fast and lightweight platform with agent mobility support. Beside the functions common to most of agent platforms it provides the Geographical Information System service to user, so it can be used for experiments with environment Simulation and communication inaccessibility \(\mathcal{A}\)-GLOBE. performance benchmarks compared against other agent platforms are also stated in this paper.
-
simulating agents mobility and inaccessibility with a globe multi agent system
Lecture Notes in Computer Science, 2004Co-Authors: David Sislak, Martin Rehak, Milan Rollo, Michal Pechoucek, Jan TožičkaAbstract:People use multi-agent systems for three kinds of activities - (i) integration (ii) Simulation (iii) complex problem solving. In this contribution we describe architecture of a newly developed multi-agent platform that supports real World Simulation, with particular emphasis to all kinds of mobility and communication inaccessibility. A-GLOBE performance benchmarks compared against other agent platforms are also stated in this paper.
Peter Wolf - One of the best experts on this subject based on the ideXlab platform.
-
learning how to drive in a real World Simulation with deep q networks
IEEE Intelligent Vehicles Symposium, 2017Co-Authors: Peter Wolf, Christian Hubschneider, Michael Weber, Andre Bauer, Jonathan Hartl, Fabian Durr, Marius J ZollnerAbstract:We present a reinforcement learning approach using Deep Q-Networks to steer a vehicle in a 3D physics Simulation. Relying solely on camera image input the approach directly learns steering the vehicle in an end-to-end manner. The system is able to learn human driving behavior without the need of any labeled training data. An action-based reward function is proposed, which is motivated by a potential use in real World reinforcement learning scenarios. Compared to a naive distance-based reward function, it improves the overall driving behavior of the vehicle agent. The agent is even able to reach comparable to human driving performance on a previously unseen track in our Simulation environment.
Jan Tožička - One of the best experts on this subject based on the ideXlab platform.
-
simulating agents mobility and inaccessibility with a globe multi agent system
Lecture Notes in Computer Science, 2004Co-Authors: David Sislak, Martin Rehak, Milan Rollo, Michal Pechoucek, Jan TožičkaAbstract:People use multi-agent systems for three kinds of activities - (i) integration (ii) Simulation (iii) complex problem solving. In this contribution we describe architecture of a newly developed multi-agent platform that supports real World Simulation, with particular emphasis to all kinds of mobility and communication inaccessibility. A-GLOBE performance benchmarks compared against other agent platforms are also stated in this paper.
Milan Rollo - One of the best experts on this subject based on the ideXlab platform.
-
a globe agent development platform with inaccessibility and mobility support
2005Co-Authors: David Sislak, Michal Pěchoucek, Martin Rehak, Milan Rollo, Dusan PavlicekAbstract:At present several Java-based multi-agent platforms from different developers are available, but none of them fully supports agent mobility and communication inaccessibility Simulation. They are thus unsuitable for experiments with large scale real-World Simulation. In this chapter we describe architecture of A-globe, fast, scalable and lightweight agent development platform with environmental Simulation and mobility support. Beside the functions common to most agent platforms it provides a position-based messaging service, so it can be used for experiments with extensive environment Simulation and communication inaccessibility. Simple benchmarks that compare the A-globe performance against other available agent platforms are also included.
-
a globe agent platform with inaccessibility and mobility support
Cooperative Information Agents, 2004Co-Authors: David Sislak, Milan Rollo, Michal PěchoucekAbstract:At present several Java-based multi-agent platforms from different developers are available, but none of them fully supports agent mobility and communication inaccessibility. They are thus no suitable for experiments with real-World Simulation. In this paper we describe architecture of newly developed agent platform \(\mathcal{A}\)-GLOBE. is fast and lightweight platform with agent mobility support. Beside the functions common to most of agent platforms it provides the Geographical Information System service to user, so it can be used for experiments with environment Simulation and communication inaccessibility \(\mathcal{A}\)-GLOBE. performance benchmarks compared against other agent platforms are also stated in this paper.
-
simulating agents mobility and inaccessibility with a globe multi agent system
Lecture Notes in Computer Science, 2004Co-Authors: David Sislak, Martin Rehak, Milan Rollo, Michal Pechoucek, Jan TožičkaAbstract:People use multi-agent systems for three kinds of activities - (i) integration (ii) Simulation (iii) complex problem solving. In this contribution we describe architecture of a newly developed multi-agent platform that supports real World Simulation, with particular emphasis to all kinds of mobility and communication inaccessibility. A-GLOBE performance benchmarks compared against other agent platforms are also stated in this paper.