Papers
8
Total Citations
171
H-Index
6
About
Giampiero Salvi is a researcher whose work sits at the intersection of robotics, cognitive systems, and language acquisition, with a particular focus on enabling robots to perceive and interact with the world in human-like ways. His most influential contributions center on multimodal learning — the integration of audio, visual, and sensorimotor information — to help robots understand and respond to their environments. Salvi's foundational work on affordance-based word-to-meaning association (2009, 30 citations) established a framework for grounding language in robotic action and perception, which he extended into a compelling model of language bootstrapping (2011, 38 citations) that mirrors how human infants acquire meaning from interaction. His research on audio-visual detection of manipulation actions (2014, 45 citations) represents his most widely recognized contribution, demonstrating that multimodal fusion significantly improves action classification in robotic systems. Beyond perception, Salvi has tackled anticipatory robot behavior through gesture recognition (2013, 23 citations) and more recently explored self-supervised active speaker detection to support socially aware language acquisition (2019, 18 citations). Across his career, his work consistently advances the goal of building robots capable of learning language and affordances through embodied, social experience.
Research Focus
Key Achievements
Top Papers
- 1Audio-visual classification and detection of human manipulation actions45 citations · 2014
- 2
- 3Affordance based word-to-meaning association30 citations · 2009
- 4
- 5
- 6
- 7Interactive Robot Learning of Gestures, Language and Affordances6 citations · 2017
- 8A dataset of human manipulation actions2 citations · 2014