Abstract
Agent-based modeling is a paradigm of modeling dynamic systems of interacting agents that are individually governed by specified behavioral rules. Training a model of such agents to produce an emergent behavior by specification of the emergent (as opposed to agent) behavior is easier from a demonstration perspective. While many approaches involve manual behavior specification via code or reliance on a defined taxonomy of possible behaviors, the meta-modeling framework in Miner [2010] generates mapping functions between agent-level parameters and swarm-level parameters, which are re-usable once generated. This work builds on that framework by integrating demonstration by image or video. The demonstrator specifies spatial motion of the agents over time and retrieves agent-level parameters required to execute that motion. The framework, at its core, uses computationally cheap image-processing algorithms. Our work is tested with a combination of primitive visual feature extraction methods (contour area and shape) and features generated using a pre-trained deep neural network in different stages of image featurization. The framework is also evaluated for its potential using complex visual features for all image featurization stages. Experimental results show significant coherence between demonstrated behavior and predicted behavior based on estimated agent-level parameters specific to the spatial arrangement of agents.
- Manuele Brambilla, Arne Brutschy, Marco Dorigo, and Mauro Birattari. 2015. Property-driven design for robot swarms: A design method based on prescriptive modeling and model checking. ACM Trans. Auton. Adapt. Syst. (TAAS) 9, 4 (2015), 17. Google Scholar
Digital Library
- Manuele Brambilla, Eliseo Ferrante, Mauro Birattari, and Marco Dorigo. 2013. Swarm robotics: A review from the swarm engineering perspective. Swarm Intel. 7, 1 (2013), 1--41.Google Scholar
Cross Ref
- Manuele Brambilla, Carlo Pinciroli, Mauro Birattari, and Marco Dorigo. 2012. Property-driven design for swarm robotics. In Proceedings of the 11th International Conference on Autonomous Agents and Multiagent Systems—Volume 1. International Foundation for Autonomous Agents and Multiagent Systems, 139--146. Google Scholar
Digital Library
- Karan K. Budhraja. 2015. Neuroevolution-Based Inverse Reinforcement Learning. Master’s thesis. University of Maryland, Baltimore County.Google Scholar
- Karan K. Budhraja and Tim Oates. 2016a. Controlling swarms by visual demonstration. In Proceedings of the 2016 IEEE 10th International Conference on Self-Adaptive and Self-Organizing Systems (SASO’16), IEEE, 1--10.Google Scholar
- Karan K. Budhraja and Tim Oates. 2016b. Neuroevolution-based inverse reinforcement learning. arXiv:1608.02971 (2016).Google Scholar
- Gilles Coppin and François Legras. 2012. Controlling swarms of unmanned vehicles through user-centered commands. In Proceedings of the AAAI Fall Symposium on Human Control of Bioinspired Swarms. 21--25.Google Scholar
- David B. D’Ambrosio, Joel Lehman, Sebastian Risi, and Kenneth O. Stanley. 2010. Evolving policy geometry for scalable multiagent learning. In Proceedings of the 9th International Conference on Autonomous Agents and Multiagent Systems: Volume 1. International Foundation for Autonomous Agents and Multiagent Systems, 731--738. Google Scholar
Digital Library
- Boris Delaunay. 1934. Sur la sphere vide. Izv. Akad. Nauk SSSR, Otdelenie Matematicheskii i Estestvennyka Nauk 7, 793-800 (1934), 1--2.Google Scholar
- Yancy Diaz-Mercado, Sung G. Lee, and Magnus Egerstedt. 2015. Distributed dynamic density coverage for human-swarm interactions. In Proceedings of the American Control Conference (ACC’15), IEEE, 353--358.Google Scholar
Cross Ref
- Jeffrey Donahue, Lisa Anne Hendricks, Sergio Guadarrama, Marcus Rohrbach, Subhashini Venugopalan, Kate Saenko, and Trevor Darrell. 2015. Long-term recurrent convolutional networks for visual recognition and description. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR’15).Google Scholar
Cross Ref
- Gianpiero Francesca, Manuele Brambilla, Arne Brutschy, Vito Trianni, and Mauro Birattari. 2014. AutoMoDe: A novel approach to the automatic design of control software for robot swarms. Swarm Intel. 8, 2 (2014), 89--112.Google Scholar
Cross Ref
- David Freelan, Drew Wicke, Keith Sullivan, and Sean Luke. 2014. Towards rapid multi-robot learning from demonstration at the robocup competition. In RoboCup 2014: Robot World Cup XVIII. Springer, 369--382.Google Scholar
- Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2015. Deep residual learning for image recognition. arXiv:1512.03385 (2015).Google Scholar
- Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural Computat. 9, 8 (1997), 1735--1780. Google Scholar
Digital Library
- Diederik P. Kingma and Max Welling. 2013. Auto-encoding variational bayes. arXiv:1312.6114 (2013).Google Scholar
- Joseph Paul Lancaster Jr. 2015. Predicting the Behavior of Robotic Swarms in Discrete Simulation. Ph.D. Dissertation. Kansas State University.Google Scholar
- Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner. 1998. Gradient-based learning applied to document recognition. Proc. IEEE 86, 11 (1998), 2278--2324.Google Scholar
Cross Ref
- Mieke Massink, Manuele Brambilla, Diego Latella, Marco Dorigo, and Mauro Birattari. 2013. On the use of Bio-PEPA for modelling and analysing collective behaviours in swarm robotics. Swarm Intel. 7, 2--3 (2013), 201--228.Google Scholar
Cross Ref
- Don Miner. 2009. Rule abstraction: Understanding emergent behavior in swarm systems. PhD Proposal, University of Maryland, Baltimore County.Google Scholar
- Donald Miner. 2010. A Framework for Predicting and Controlling System-Level Properties of Agent-Based Models. Ph.D. Dissertation. University of Maryland, Baltimore County. Google Scholar
Digital Library
- Don Miner and others. 2009. Predicting and controlling system-level parameters of multi-agent systems. In Proceedings of the 2009 AAAI Fall Symposium Series.Google Scholar
- Don Miner and Marie desJardins. 2008. Learning abstract properties of swarm systems. In Proceedings of the 8th International Conference on Autonomous Agents and Multiagent Systems. Citeseer.Google Scholar
- Don Miner and Marie desJardins. 2009. Learning non-explicit control parameters of self-organizing systems. In Proceedings of the 3rd IEEE International Conference on Self-Adaptive and Self-Organizing Systems (SASO’09).. IEEE, 286--287. Google Scholar
Digital Library
- Don Miner, Marie desJardins, and Peter Hamilton. 2008. The swarm application framework. In Proceedings of the 23rd National Conference on Artificial Intelligence—Volume 3. AAAI Press, 1822--1823. Google Scholar
Digital Library
- Timo Ojala, Matti Pietikainen, and David Harwood. 1994. Performance evaluation of texture measures with classification based on Kullback discrimination of distributions. In Proceedings of the 12th IAPR International Conference on Pattern Recognition, 1994. Vol. 1—Conference A: Computer Vision & Image Processing, Vol. 1. IEEE, 582--585.Google Scholar
Cross Ref
- Timo Ojala, Matti Pietikäinen, and David Harwood. 1996. A comparative study of texture measures with classification based on featured distributions. Pattern Recogn. 29, 1 (1996), 51--59.Google Scholar
Cross Ref
- Gabriele Paolacci, Jesse Chandler, and Panagiotis G. Ipeirotis. 2010. Running experiments on amazon mechanical turk. Judg. Dec. Mak. 5, 5 (2010), 411--419.Google Scholar
- Danilo Pianini, Mirko Viroli, and Jacob Beal. 2015. Protelis: Practical aggregate programming. In Proceedings of the 30th Annual ACM Symposium on Applied Computing. ACM, 1846--1853. Google Scholar
Digital Library
- Daniel Pickem, Li Wang, Paul Glotfelter, Yancy Diaz-Mercado, Mark Mote, Aaron Ames, Eric Feron, and Magnus Egerstedt. 2016. Safe, remote-access swarm robotics research on the robotarium. arXiv:1604.00640 (2016).Google Scholar
- John Reeder. 2015. Team search tactics through multi-agent HyperNEAT. In Information Processing in Cells and Tissues. Springer, 75--89.Google Scholar
- Karen Simonyan and Andrew Zisserman. 2014. Very deep convolutional networks for large-scale image recognition. arXiv:1409.1556 (2014).Google Scholar
- Adrian Šošić, Wasiur R. KhudaBukhsh, Abdelhak M. Zoubir, and Heinz Koeppl. 2016. Inverse reinforcement learning in swarm systems. arXiv:1602.05450 (2016).Google Scholar
- Keith Sullivan and Sean Luke. 2012. Learning from demonstration with swarm hierarchies. In Proceedings of the 11th International Conference on Autonomous Agents and Multiagent Systems—Volume 1. International Foundation for Autonomous Agents and Multiagent Systems, 197--204. Google Scholar
Digital Library
- Andreas Veit, Michael Wilber, and Serge Belongie. 2016. Residual networks are exponential ensembles of relatively shallow networks. arXiv:1605.06431 (2016).Google Scholar
- Uri Wilensky. 1998. NetLogo flocking model. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL.Google Scholar
- Uri Wilensky and I. Evanston. 1999. NetLogo: Center for connected learning and computer-based modeling. Northwestern University, Evanston, IL (1999), 49--52.Google Scholar
- Kevin Winner, Don Miner, and Marie desJardins. 2009. Controlling particle swarm optimization with learned parameters. In Proceedings of the 3rd IEEE International Conference on Self-Adaptive and Self-Organizing Systems (SASO’09). IEEE, 288--290. Google Scholar
Digital Library
- SHI Xingjian, Zhourong Chen, Hao Wang, Dit-Yan Yeung, Wai-kin Wong, and Wang-chun Woo. 2015. Convolutional LSTM network: A machine learning approach for precipitation nowcasting. In Proceedings of the Conference on Advances in Neural Information Processing Systems. 802--810. Google Scholar
Digital Library
Index Terms
Feature Construction for Controlling Swarms by Visual Demonstration
Recommendations
Physical parameter optimization in swarms of ultra-low complexity agents
AAMAS '08: Proceedings of the 7th international joint conference on Autonomous agents and multiagent systems - Volume 3Physical agents (such as wheeled vehicles, UAVs, hovercraft, etc.) with simple control systems are often sensitive to changes in their physical design and control parameters. As such, it is crucial to evaluate the agent's control systems together with ...
Polymorphic self-* agents for stigmergic fault mitigation in large-scale real-time embedded systems
AAMAS '05: Proceedings of the fourth international joint conference on Autonomous agents and multiagent systemsOrganization and coordination of agents within large-scale, complex, distributed environments is one of the primary challenges in the field of multi-agent systems. A lot of interest has surfaced recently around self-* (self-organizing, self-managing, ...
Learning from demonstration with swarm hierarchies
AAMAS '12: Proceedings of the 11th International Conference on Autonomous Agents and Multiagent Systems - Volume 1We present a supervised learning from demonstration system capable of training stateful and recurrent collective behaviors for multiple agents or robots. A model space of this kind is often high-dimensional and consequently may require a large number of ...






Comments