Vous êtes ici : GIPSA-lab >CRISSPHome CRISSP
 
Team

COGNITIVE ROBOTICS, INTERACTIVE SYSTEMS, & SPEECH PROCESSING
Team manager : Gérard BAILLYThomas HUEBER

 

CRISSP team conducts theoretical, experimental and technological researches in the field of speech communication. More precisely, we aim at: 

    • Modeling verbal and co-verbal speech signal in face-to-face interaction involving humans, virtual avatar (talking head) and humanoid robots.
    • Understanding the human speech production process by modeling relationships between speech articulation and speech acoustics.
    • Studying communication of people with hearing impairment.
    • Designing speech technologies for handicapped people, language learning, and multimedia.

 

 

 

The 3 research axis of the CRISSP team are:

    • Cognitive robotics: improve socio-communicative skills of humanoid robots. 
    • Interactive systems: design real-time/reactive communicative systems exploiting the different modalities of speech (audio, visual, gesture, etc.).
    • Speech processing: articulatory synthesis, acoustic-articulatory inversion, speech synthesis, voice conversion.

Domains of expertise of CRISSP team

    • Audio signal processing (analysis, coding, denoising, source separation)
    • Speech processing (analysis, transformation, conversion/morphing, text-to-speech synthesis, articulatory synthesis/inversion)
    • Statistical machine learning
    • Acquisition of multimodal articulatory data (using electromagnetic articulography, ultrasound imaging, MRI, EMG, etc.)
    • Acquisition of social signals (eye gaze, body posture, head movements, etc.) during face-to-face interaction

 

Team members

(updated 18/12/2015)

 

Contact : Gérard Bailly et Thomas Hueber (mail : firstname.lastname@gipsa-lab.fr)



News
RecrutementGIPSA-Lab recrute pour 12 mois un Post-doc (H/F) - Evaluation automatique de la fluence de jeunes lecteurs, pour l'équipe CRISSP

La caractéristique importante des lectures à haute voix produites par les lecteurs débutants est la prévalence de disfluences telles que pauses incongrues, répétitions et faux départs. L'objectif de ce travail est de détecter ces disfluences et de quantifier leur impact sur la qualité perçue de la lecture en utilisant les techniques d'apprentissage automatique à l'état de l'art. Dans le cadre du projet e-FRAN Fluence (fluence.prod.lamp.cnrs.fr), nous allons en effet collecter des lectures enfantines de manière massive.

Lire la suite



Last publications of team

Semi-supervised multichannel speech enhancement with variational autoencoders and non-negative matrix factorization

Simon Leglaive, Laurent Girin, Radu Horaud. Semi-supervised multichannel speech enhancement with variational autoencoders and non-negative matrix factorization. IEEE International Conference on Acoustics Speech and Signal Processing (ICASSP 2019), May 2019, Brighton, United Kingdom. IEEE, pp.1-5. 〈hal-02005102〉

Speech enhancement with variational autoencoders and alpha-stable distributions

Simon Leglaive, Umut Simsekli, Antoine Liutkus, Laurent Girin, Radu Horaud. Speech enhancement with variational autoencoders and alpha-stable distributions. ICASSP 2019 - IEEE International Conference on Acoustics Speech and Signal Processing, May 2019, Brighton, United Kingdom. IEEE, pp.1-5, 2019, 〈https://2019.ieeeicassp.org〉. 〈hal-02005106〉

Multichannel Speech Separation and Enhancement Using the Convolutive Transfer Function

Xiaofei Li, Laurent Girin, Sharon Gannot, Radu Horaud. Multichannel Speech Separation and Enhancement Using the Convolutive Transfer Function. IEEE/ACM Transactions on Audio, Speech and Language Processing, Institute of Electrical and Electronics Engineers, 2019, 27 (3), pp.645-659. 〈10.1109/TASLP.2019.2892412〉. 〈hal-01799809〉


Voir toutes les publications de l'équipe dans HAL
GIPSA-lab, 11 rue des Mathématiques, Grenoble Campus BP46, F-38402 SAINT MARTIN D'HERES CEDEX - 33 (0)4 76 82 71 31