Pauline Lavalade

Sorbonne Université, University of Plymouth

Papers

Total Citations

161

H-Index

About

Pauline Lavalade is a leading researcher at the intersection of child–robot interaction and speech technology. Her work addresses a critical bottleneck in human-robot interaction (HRI): the reliable recognition of children’s speech. While automatic speech recognition (ASR) has advanced rapidly for adults, children’s variable articulation, pitch, and prosody pose unique challenges—challenges Lavalade has systematically tackled. Her most cited work, “Child Speech Recognition in Human-Robot Interaction” (2017, 158 citations), is a foundational study that benchmarks ASR performance with child speakers in real-world HRI settings, revealing the limitations of off-the-shelf systems and proposing tailored acoustic and language models. To support this research, she created and publicly released the “Children Speech Recording” dataset (2016), a carefully curated corpus of spontaneous and prompted English speech from 11 children (mean age 4.9 years). This resource has become a vital tool for the community, enabling reproducible work on child ASR. Lavalade’s contributions are essential for deploying robots in educational, therapeutic, and caregiving contexts, where reliable verbal interaction with young users is paramount. Her work continues to shape how robots listen to, and learn from, children.