Partners: GIPSA-Lab, LPNC
(S. Valdois, coordinatrice), LIDILEM (M. Masperi)
, LANSAD
(N. Chalon)
Objectives: evaluate computerized reading assistance for
improving fluency of young readers. GIPSA-Lab tests the RAKE (Reading
Assistance by Karaoke) system that enlightens parts of text (syllables,
words, breath groups) as they are read by an expert reader. The ELARGIR
method consists in repeated Reading-while-Listening (RwL) tasks using
the RAKE technology paced by evaluation by pairs.
Collaborations: Rectorat Grenoble (M. Zanoni,
C. Lequette, J. Eudes), LSE
(M. Bianco)
Partners: GIPSA-Lab (G. Bailly, coordinateur), Lab-STICC (P. De
Loor), LIG
(H. Fiorino), LIP
(M. Dubois), Aldebaran
Robotics (R. Gelin)
Objectives: proposes to train humanoid robots (Nina and
Roméo) i.e. provide them with sociocommunicative
abilities by immersive teleoperation i.e.
“beaming” of human pilots. The SOMBRERO
project has three main objectives:
Realize beaming experiments involving adults
interacting with a humanoid robot in cooperative, situated and
finalized tasks. The targeted task is a series of tests used in
neuropsychological records. We focus on the socio-communicative
behavior that should accompany the monitoring of the task, i.e. the
correct comprehension of the instructions, the seamless execution of
the task by the interlocutor, the positive feedback and encouragement
given along the performance, correction of errors, misunderstandings or
precisions,
Develop and implement autonomous socio-communicative
behaviors into the robot cognitive architecture via statistical
modeling of the multimodal behaviors monitored during the prior
robotmediated interactions,
Assess these behaviors and the achieved social
embodiment with acceptance measures and analysis of user attitudes.
Partners: GIPSA-Lab, LPNC
(S. Valdois, coordinatrice), LpnCog
(S. Pacton), IUHC (M. Zorman) & Play
Bac Editions (E.
Serreno-Despres)
Objectives: role for the visual attention span in
orthographic processing and orthographic learning. GIPSA will implement
and study the impact of synchronous reading/listening of audio books on
the acquisition of orthography.
Resources: audiobooks for Windows (each archive includes
one directory
with all ressources for synchronous reading):
Jules Vernes (Tour du monde :
chapitres 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
)
Objectives: this
project studies decisional and operational human-robot
interaction, and more specifically, the impact of verbal and non-verbal
communication on the execution of collaborative tasks between a robot
and a human partner.
Objectives: this
project studies silent speech interfaces, i.e.
computer-assisted
communication with silent speech production using various technologies
such as ultrasound imaging, electro-articulography, electro-myography
and stethoscopic microphones.
Objectives: Acoustic-to-articulatory
inversion for speech
pronunciation training. GIPSA-Lab promotes a data-driven approach with
direct (GMM-based) versus language-specific (HMM-based) mappings.
ARTUS
(RNRT 2002-2005)
Partners: LIS (J.-M. Chassery &
P. Bas,
coordinateurs), EUDIASYC (F. Davoine), TSI/ENST (N. Moreau), ARTE Labs
(J.P. Léoni), Attitude Studio (R. Brun) & Nextamp
(P. Nguyen)
Objectives: Virtual speech cuer as an
alternative for
teletext for deaf televiewers. Gestures are watermarked in the
audiovisual flow.
MOTHER
(ICP 2000)& VESALE
(BQR INP 2003-2004)
Partners: TU Berlin (PHC PROCOPE
2006-2007 "static and
dynamic
replication of talking faces" with S. Fagel) & FT R&D
(contract 2005-2007 "Animation de visages parlants" with G. Breton)
Objectives: Virtual clones of speakers
Resources:
Original speakers: SF
(German), HL
(French), CD (Australian), OC (English) & PDB (English, model)