**************** LPCSR project code *************************
Speech analysis and synthesis using linear predictive coding (LPC) in Matlab. Also includes a naive speech recognition script using LPC functions.
LPCSR_AddOverlap.m function: combine synthesis output frames
LPCSR_Analysis.m function: estimate LPC parameters from input audio frame ("transmitter")
LPCSR_Autocorr.m function: estimate pitch and decide whether a frame is voiced or unvoiced
LPCSR_DecodeParams.m function: convert line spectral pair (LSP) parameter representation to LPC coefficients
LPCSR_EncodeParams.m function: convert LPC coeffs to LSP format
LPCSR_ExampleRecognition.m script: example of simple speech recognition based on LPC analysis of input audio
LPCSR_ExampleSynthesis.m script: example of analysis and resynthesis of an input audio file using LPC
LPCSR_Excite.m function: generate excitation frame from pitch and voiced/unvoiced parameters
LPCSR_LoopWrapper.m function: recordkeeping to handle iteration across an input file
LPCSR_NoisyTransmit.m function: simple simulation of noisy transmission channel from "transmitter" to "receiver"
LPCSR_PlotAnalysis.m function: generate plots of input, analysis, synthesis, and output data
LPCSR_Prefilter.m function: pre-analysis filtering of input audio
LPCSR_ProcessSamples.m function: calculate LPC coefficients for samples in a database
LPCSR_Prony.m function: estimate LPC filter coefficients and gain from an input frame of audio data
LPCSR_Recognize.m function: wrapper to perform speech recognition
LPCSR_Synthesis.m function: synthesize speech based on input LPC parameters ("receiver")
README.txt this file