7th International Conference on Spoken Language Processing
September 16-20, 2002
The Early Vocalization Analyzer (EVA) is a computer program that automatically analyses digitized acoustic recordings of infant vocalizations. Using the landmark detection theory of Stevens et al for the recognition of phonetic features in speech, EVA detects syllables in vocalizations produced by infants. Landmarks are grouped into standard syllable patterns and syllables are grouped into utterances. Statistics derived from these groups and the underlying features are used to derive a "vocalization age" and two specific screening rules that can clinically distinguish infants who may be at risk for later communication or other developmental problems from typically developing infants in the six to fifteen month age range.
Bibliographic reference. Fell, Harriet J. / MacAuslan, Joel / Ferrier, Linda J. / Worst, Susan G. / Chenausky, Karen (2002): "Vocalization age as a clinical tool", In ICSLP-2002, 2345-2348.