Compute the GeMAPS openSMILE feature set (C++ Implementation)

This function applies the "The Geneva Minimalistic Acoustic Parameter Set (GeMAPS) for Voice Research and Affective Computing" (the Geneva Minimalistic Standard Parameter Set, GeMAPS v0.1b) (Eyben et al. 2015) to a portion of a recording using the native OpenSMILE C++ library for maximum performance.

Usage

lst_GeMAPS(listOfFiles, beginTime = 0, endTime = 0, explicitExt = "gem", verbose = FALSE, toFile = FALSE, return_jstf = FALSE, outputDirectory = NULL)

Arguments

listOfFiles

The full path to the sound file.

beginTime

The starting time of the section of the sound files that should be analysed (in seconds).

endTime

The end time of the section of the sound files that should be analysed (in seconds). Use 0 for end of file.

explicitExt

The file extension of the slice file where the results should be stored. Default "gem".

verbose

Logical. Print processing information (default: FALSE).

toFile

Logical. If TRUE, write results to JSTF file. Default FALSE.

return_jstf

Logical. Return JsonTrackObj instead of list? Default FALSE. When both toFile and return_jstf are TRUE, writes the file AND returns the object.

outputDirectory

Character. Output directory path. Default NULL (use input directory).

Value

If return_jstf=FALSE and toFile=FALSE (default), a named list of 62 acoustic values. If toFile=TRUE, invisibly returns the path(s) to the written JSTF file(s). If return_jstf=TRUE, returns a JsonTrackObj.

Details

The GeMAPS feature set consists of 62 static acoustic features (pitch, jitter, shimmer, formants, HNR, spectral balance, and loudness) computed via the openSMILE C++ library directly (3-5x faster than the Python implementation). See the openSMILE GeMAPS paper ((Eyben et al. 2015) ) for the full definition of each descriptor and functional.

References

Eyben F, Scherer KR, Schuller BW, Sundberg J, Andre E, Busso C, Devillers LY, Epps J, Laukka P, Narayanan SS, Truong KP (2015). “The Geneva Minimalistic Acoustic Parameter Set (GeMAPS) for Voice Research and Affective Computing.” IEEE Transactions on Affective Computing, 7(2), 190–202. ISSN 1949-3045. doi:10.1109/taffc.2015.2457417 .

Examples

if (FALSE) { # \dontrun{
# Using C++ implementation (default, fastest)
gemaps <- lst_GeMAPS("audio.wav")

# With time windowing
gemaps <- lst_GeMAPS("audio.wav", beginTime = 1.0, endTime = 3.0)

# Write results to JSTF file
lst_GeMAPS("audio.wav", toFile = TRUE)  # Creates audio.gem

# Read back and convert to data.frame
track <- read_track("audio.gem")
df <- as.data.frame(track)
head(df)  # Shows begin_time, end_time, and all 62 GeMAPS features
} # }