Pierre Yves Mignotte

Papers

Total Citations

H-Index

About

Pierre Yves Mignotte is a researcher at the intersection of natural language processing, computer vision, and human-robot interaction. His primary focus lies in developing multimodal systems that enable more intuitive communication between humans and machines, particularly through the integration of visual and linguistic data. Mignotte’s most notable contribution is the creation of the ROSMI (Robot Open Street Map Instructions) corpus, a pioneering multimodal dataset comprising map-based instruction pairs collected via crowdsourcing. This publicly available resource, introduced in 2020, is designed to advance state-of-the-art visual-dialogue tasks by providing rich, real-world examples of how spatial information is conveyed through natural language. While his work has garnered modest citation counts to date, the ROSMI corpus represents a foundational step toward improving robots’ ability to understand and follow human instructions in complex environments. Mignotte’s research is particularly valuable for students and researchers exploring grounded language understanding, situated dialogue, and the challenges of deploying autonomous systems in dynamic, map-based settings.