CNRS Research Director
Research topics: Modelling and Automatic Processing of Sign Language, from corpus linguistics to artificial intelligence.
Research questions :
How can sign languages be analysed, represented and processed ?
How can we take into account the specific linguistic characteristics linked to their visual-gestural nature (multilinearity, spatialisation, iconicity)?
What types of approach are possible with little French sign language (LSF) data?
Research work :
Corpus of LSF: constitution, annotation, analysis
Nature of the data: video, motion capture (mocap)
Multidisciplinary studies: computer science (analysis and synthesis), linguistics, movement sciences, movement perception
Linguistic modelling to represent the functioning of LSF
Levels: articulatory, gestural unit, more complex structures
Articulators: hands/arms, bust, shoulders, head, face, facial elements (eyebrows, eyes, cheeks, mouth), gaze
Animation of virtual signers and generation of speech in LSF
Creation of animations by rotoscopy, mocap or synthesis
Generation of statements from predefined animations
Analysis of LSF videos
Annotation assistance
Automatic recognition
Publications sur HAL - Corpora sur Ortolang.
Read more…