Sorin Moga

About

Sorin Moga is a pioneering researcher in autonomous robotics and computational neuroscience, whose work has fundamentally shaped our understanding of machine learning through imitation. Working primarily from the late 1990s through the early 2000s, Moga developed innovative neural network architectures enabling robots to learn by observing and imitating human or robotic movement — a capability once thought to require explicit programming. His most celebrated contribution, exploring how imitation can emerge organically from "perception ambiguity" rather than being hard-coded into a system, garnered over 130 citations and established a bottom-up paradigm for robotic learning. His follow-up work connecting imitation to communication and turn-taking behavior accumulated nearly 120 citations, demonstrating that gestural proto-communication could arise naturally from these same neural principles. Beyond imitation, Moga explored dynamical neural field models for robot navigation and path planning, hippocampal-inspired sequence learning architectures, and parallel computing approaches for neural simulation. His research bridges developmental psychology, neuroscience, and robotics, drawing meaningful inspiration from biological systems to solve real-world challenges in autonomous learning. With over 360 total citations, Moga's body of work remains a foundational reference for researchers in embodied cognition and developmental robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   FROM PERCEPTION-ACTION LOOPS TO IMITATION PROCESSES: A BOTTOM-UP APPROACH OF LEARNING BY IMITATION

   133 citations · 1998
2. 2
   Learning and communication via imitation: an autonomous robot perspective

   118 citations · 2001
3. 3
   Dynamical neural networks for planning and low-level robot control

   40 citations · 2003
4. 4
   From Perception-Action loops to imitation processes: A bottom-up approach of learning by imitation

   15 citations · 1997
5. 5
   Apprendre par imitation : une nouvelle voie d'apprentissage pour les robots autonomes

   15 citations · 2000
6. 6
   Imitation: Learning and communication

   14 citations · 2000
7. 7
   A Neural Structure for Learning by Imitation

   10 citations · 1999
8. 8
   Sequence learning and timing in hippocampus, prefrontal cortex, and accumbens

   7 citations · 2002
9. 9
   Autonomous robot learning: what can we take for free?

   6 citations · 2002
10. 10
    Parallelization of Neural Networks Using PVM

    5 citations · 2000

Key Collaborators

Contact & Links