Document Detail


Data-driven automated acoustic analysis of human infant vocalizations using neural network tools.
MedLine Citation:
PMID:  20370038     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
Acoustic analysis of infant vocalizations has typically employed traditional acoustic measures drawn from adult speech acoustics, such as f(0), duration, formant frequencies, amplitude, and pitch perturbation. Here an alternative and complementary method is proposed in which data-derived spectrographic features are central. 1-s-long spectrograms of vocalizations produced by six infants recorded longitudinally between ages 3 and 11 months are analyzed using a neural network consisting of a self-organizing map and a single-layer perceptron. The self-organizing map acquires a set of holistic, data-derived spectrographic receptive fields. The single-layer perceptron receives self-organizing map activations as input and is trained to classify utterances into prelinguistic phonatory categories (squeal, vocant, or growl), identify the ages at which they were produced, and identify the individuals who produced them. Classification performance was significantly better than chance for all three classification tasks. Performance is compared to another popular architecture, the fully supervised multilayer perceptron. In addition, the network's weights and patterns of activation are explored from several angles, for example, through traditional acoustic measurements of the network's receptive fields. Results support the use of this and related tools for deriving holistic acoustic features directly from infant vocalization data and for the automatic classification of infant vocalizations.
Authors:
Anne S Warlaumont; D Kimbrough Oller; Eugene H Buder; Rick Dale; Robert Kozma
Related Documents :
19710708 - Convergent temporal dynamics of the human infant gut microbiota.
18758718 - Control pattern of vocal center for vocalization in ruddy bunting (emberiza rutila).
20589708 - Time windows in retention over the first year-and-a-half of life: spacing effects.
18973908 - Causal perception of action-and-reaction sequences in 8- to 10-month-olds.
15173488 - Serious bacterial infections in febrile infants 1 to 90 days old with and without viral...
22112238 - Peritoneal drainage versus laparotomy for perforated necrotising enterocolitis or spont...
Publication Detail:
Type:  Comparative Study; Journal Article; Research Support, N.I.H., Extramural; Research Support, U.S. Gov't, Non-P.H.S.    
Journal Detail:
Title:  The Journal of the Acoustical Society of America     Volume:  127     ISSN:  1520-8524     ISO Abbreviation:  J. Acoust. Soc. Am.     Publication Date:  2010 Apr 
Date Detail:
Created Date:  2010-04-07     Completed Date:  2010-07-06     Revised Date:  2011-07-27    
Medline Journal Info:
Nlm Unique ID:  7503051     Medline TA:  J Acoust Soc Am     Country:  United States    
Other Details:
Languages:  eng     Pagination:  2563-77     Citation Subset:  IM    
Affiliation:
School of Audiology and Speech-Language Pathology, The University of Memphis, 807 Jefferson Avenue, Memphis, Tennessee 38105, USA. awarlmnt@memphis.edu
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Acoustics*
Age Factors
Algorithms*
Automation
Child Language*
Female
Humans
Infant
Male
Models, Biological*
Neural Networks (Computer)*
Phonation*
Reproducibility of Results
Signal Processing, Computer-Assisted*
Sound Spectrography
Time Factors
Voice*
Grant Support
ID/Acronym/Agency:
R01 DC006099-04/DC/NIDCD NIH HHS; R01 DC006099-04/DC/NIDCD NIH HHS; R01 DC006099-05/DC/NIDCD NIH HHS
Comments/Corrections

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Dependence of phonation threshold pressure and frequency on vocal fold geometry and biomechanics.
Next Document:  Acoustic characteristics of phonation in "wet voice" conditions.