Functions for applying manifold-and-eigenmapping methods to high-dimensional speech data to learn low-dimensional representations of speech for phonetic analysis. Includes a data set that comprises productions of the sibilant fricatives /s, S/ by native American English-speaking adults, productions by native American English-acquiring two- and three-year-old children, and ratings of the children's productions by native American English-speaking adult listeners. Includes a vignette on how to construct a manifold from speech data and then eigenmap it to a low-dimensional representation on which phonetic analysis can proceed. Accompanies the manuscript: A. R. Plummer and P. F. Reidy (submitted). "Computing low-dimensional representations of speech from socio-auditory structures for phonetic analyses". Journal of Phonetics.
phoneigen may be installed directly from GitHub:
devtools::install_github("patrickreidy/phoneigen", build_vignettes = TRUE)After the package has been installed on your machine, you can access the
vignette that accompanies the manuscript submitted to Journal of Phonetics
by calling vignette("socio-auditory-manifolds", "phoneigen").