Human vs. Deep Neural Network Performance at a Leader Identification Task

Deka, Ankur and Sycara, Katia and Walker, Phillip and Li, Huao and Lewis, Michael (2021) Human vs. Deep Neural Network Performance at a Leader Identification Task. In: Proceedings of the 65th Annual Meeting of the Human Factors and Ergonomics Society. (In Press)

Preview

PDF Download (336kB) | Preview

Abstract

Control of robotic swarms through control over a leader(s) has become the dominant approach to supervisory control over these largely autonomous systems. Resilience in the face of attrition is one of the primary advantages attributed to swarms yet the presence of leader(s) makes them vulnerable to decapitation. Algorithms which allow a swarm to hide its leader are a promising solution. We present a novel approach in which neural networks, NNs, trained in a graph neural network, GNN, replace conventional controllers making them more amenable to training. Swarms and an adversary intent of finding the leader were trained and tested in 4 phases: 1-swarm to follow leader, 2-adversary to recognize leader, 3-swarm to hide leader from adversary, and 4-swarm and adversary compete to hide and recognize the leader. While the NN adversary was more successful in identifying leaders without deception, humans did better in conditions in which the swarm was trained to hide its leader from the NN adversary. The study illustrates difficulties likely to emerge in arms races between machine learners and the potential role humans may play in moderating them.

---

Share

Citation/Export:	Select format... Citation - Text Citation - HTML Endnote BibTex Dublin Core OpenURL MARC (ISO 2709) METS MODS EP3 XML Reference Manager Refer
Social Networking:	Share |

---

Details

Item Type:	Conference or Workshop Item (Paper)
Status:	In Press
Creators/Authors:	Creators Email Pitt Username ORCID Deka, Ankur Sycara, Katia Walker, Phillip Li, Huao Lewis, Michael cmlewis@pitt.edu cmlewis 0000-0002-1013-9482
Date:	2021
Date Type:	Publication
Journal or Publication Title:	Proceedings of the 65th Annual Meeting of the Human Factors and Ergonomics Society
Publisher:	Human Factors and Ergonomics Society
Event Title:	Proceedings of the 65th Annual Meeting of the Human Factors and Ergonomics Society
Event Type:	Conference
Schools and Programs:	School of Computing and Information > Information Science
Refereed:	Yes
Date Deposited:	13 Aug 2021 20:17
Last Modified:	13 Aug 2021 20:17
URI:	http://d-scholarship.pitt.edu/id/eprint/41664

---

Metrics

Monthly Views for the past 3 years

Plum Analytics

---

Actions (login required)

View Item