About

Eric Martinson is a robotics researcher whose work spans multi-robot coordination, robot audition, and human-robot interaction. His early contributions focused on applying reinforcement learning — particularly Q-learning — to complex robotic systems, developing scalable approaches that grouped behaviors into "roles" to manage state-space complexity in mission-tasked robot teams, work that garnered over 40 citations. Martinson made significant strides in robot audition, introducing Auditory Evidence Grids to address the persistent challenges of sound source localization in noisy real-world environments (38 citations), and developing autonomous systems capable of discovering and mapping ambient sound sources. His research consistently bridged sensing modalities: a notable 2007 paper demonstrated how vision, acoustics, and natural language could be combined to resolve ambiguities in human-robot communication (36 citations). Martinson also contributed to multi-robot localization through Intrinsic Localization and Mapping, using robots themselves as landmarks. Later work explored person identification using soft biometrics for sustained social interaction. Across more than a decade of research, his publications have collectively accumulated hundreds of citations, reflecting sustained influence in autonomous robotics, spatial awareness, and intelligent human-robot interfaces.

Research Focus

Key Achievements

13
H-Index
38
Papers
512
Total Citations
13
Avg Citations/Paper
🏆 Most Cited Paper
Learning to role-switch in multi-robot systems
41 citations · 2004
📈 Most Prolific Year: 2007 (5 Papers)
🤝 Key Collaborators: 29
🏛 Institutions: Georgia Institute of Technology, United States Naval Research Laboratory, Toyota Motor Corporation (United States), Toyota Motor Corporation (Switzerland), Lawrence Technological University, EMAG Technologies (United States)

Top Papers

  1. 1
  2. 2
    Auditory Evidence Grids
    38 citations · 2006
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago