Papers
32
Total Citations
645
H-Index
12
About
Hiroaki Arie is a pioneering researcher in neuro-robotics, cognitive computation, and human-robot interaction, whose work sits at the compelling intersection of neuroscience-inspired machine learning and autonomous robotic systems. His research has made substantial contributions to how robots perceive, learn, and interact with complex real-world environments through multimodal integration, stochastic neural network models, and dynamic behavior generation. Arie's most influential work, "Multimodal Integration Learning of Robot Behavior Using Deep Neural Networks" (2014, 196 citations), demonstrated how deep learning could enable robots to fuse sensory modalities in ways analogous to human perception — a breakthrough with significant implications for deploying robots in everyday human environments. His development of stochastic recurrent neural network architectures, particularly the S-MTRNN model, advanced the field's understanding of how robots can distinguish between deterministic and probabilistic behavioral regimes through social interaction. Beyond perception, Arie has tackled language-behavior integration, imitation learning, and adaptive human-robot collaboration, consistently grounding his computational models in neuroscientific principles such as cortical chaos and multiple timescales processing. His body of work collectively reflects a vision of robots that don't merely execute tasks but genuinely learn, adapt, and communicate — making his research essential reading for anyone exploring the future of intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Multimodal integration learning of robot behavior using deep neural networks196 citations · 2014
- 2
- 3
- 4
- 5
- 6Imitating others by composition of primitive actions: A neuro-dynamic model25 citations · 2011
- 7
- 8
- 9
- 10Reinforcement learning of a continuous motor sequence with hidden states13 citations · 2007