Laurence Vidrascu
Papers
4
Total Citations
391
H-Index
4
About
Laurence Vidrascu is a leading researcher in the field of affective computing and speech processing, with a primary focus on the automatic classification of emotional user states from speech. Her work is foundational in understanding how acoustic and linguistic features can be harnessed to detect and categorize emotions in human-computer interaction. Her most influential contribution, the 2007 paper "The relevance of feature type for the automatic classification of emotional user states," which has garnered 189 citations, systematically explored how different low-level descriptors and functionals impact classification accuracy, using a German database of children interacting with a pet robot. This work, along with her highly cited 2010 study "Whodunnit – Searching for the most important feature types signalling emotion-related user states in speech" (146 citations), established a benchmark for feature engineering in emotion recognition. Vidrascu’s research has demonstrated the critical role of prosodic features, such as F0 extraction, in distinguishing emotional states, while also questioning traditional assumptions about pitch’s dominance in prominence marking. Her contributions have significantly advanced the development of more robust, data-driven models for analyzing paralinguistic cues in speech, impacting fields from human-robot interaction to affective computing.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Patterns, prototypes, performance: classifying emotional user states41 citations · 2008
- 4