Papers
8
Total Citations
278
H-Index
7
About
Alain Droniou is a leading researcher at the intersection of developmental robotics and deep learning, whose work focuses on endowing robots with the ability to learn autonomously through active exploration and curiosity. His major contributions center on creating cognitive architectures that allow humanoid robots, particularly the iCub, to build visual representations of objects incrementally and learn action repertoires with minimal prior knowledge. His highly cited paper "Object Learning Through Active Exploration" (68 citations) introduces a developmental approach where robots actively manipulate objects to construct visual representations, while "Deep unsupervised network for multimodal perception, representation and classification" (63 citations) advances unsupervised learning for integrating multiple sensory modalities. Droniou's influential "Towards Deep Developmental Learning" (50 citations) critically examines deep learning techniques from a developmental robotics perspective, proposing frameworks for constructing hierarchical multimodal representations. His work has demonstrated practical applications in handwriting generation and curiosity-driven object recognition, with cumulative citations exceeding 270 across his most impactful publications. Droniou's research is foundational to the MACSi project, supporting experiments that enable humanoid robots to gradually expand their knowledge through interaction with objects and caregivers, making him a key figure in advancing autonomous robotic learning systems.
Research Focus
Key Achievements
Top Papers
- 1Object Learning Through Active Exploration68 citations · 2014
- 2
- 3Towards Deep Developmental Learning50 citations · 2015
- 4Learning a repertoire of actions with deep neural networks29 citations · 2014
- 5
- 6
- 7
- 8