Auditory-Visual Speech Processing (AVSP'98)
December 4-6, 1998
The aim of this research is to identify visual cues for Italian vocalic and consonantal visemes on a articulatory basis and to verify the results of visual intelligibility tests. These data will be useful when per-forming cross-linguistic comparisons, in defining some relevant visual parameters and variation ranges which could be used in the development of bimodal automatic speech synthesis and recognition systems.
Bibliographic reference. Caldognetto, Emanuela Magno / Zmarich, Claudio / Cosi, Piero (1998): "Statistical definition of visual information for Italian vowels and consonants", In AVSP-1998, 135-140.