SignalProcessing

October 11, 2014 ยท View on GitHub

Note: The functionality of this package has been absorbed by Julia DSP. I'll keep this here for historic reasons.

This module contains some signalprocessing routines necessary for feature extraction for the most common features (MFCC, PLP) in speech processing.

Most of this is re-coded from octave 3.2.4.

Windowing (tapers)

  • hamming(n::Int)
  • hamming(n::Int)

Spectogram

  • specgram(x::Vector, n::Int=256; sr=8000, window=hamming(n), overlap=n/2)
    • computes spectogram for data in x over window of size n overlapping overlap points

Miscelaneous

  • levinson(acf::Vector, p::Int)
  • toeplitz(c::Vector, r::Vector=c)