Skip to content

Here, we aim to synthesis a noise-free, artificial speech from natural speech using LPC coefficients. Spoiler alert, LPC is very bad, but at least we've figured that pole order at around 24-28 is quite optimal for the task

Notifications You must be signed in to change notification settings

5thGenDev/Speech-Synthesis-with-LPC-coefficients

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Speech-Synthesis-with-LPC-coefficients

Specification that guides the whole coursework is in Assignment1 file.

Natural speeches are wav file with short file name, the first part is the word sample, the second part is the gender that said the word.

Artifial speeches created by LPC function have different parameters -- (1) AR:Autoregression pole order, -- (2) Segment length: we segment speech to short segments at different starting points to see how it might affect speech synthesis quality -- (3) Impulse waveform: conventional wisdom says that triangle wave should bring the most quality synthesis. Well that's not the case here!

About

Here, we aim to synthesis a noise-free, artificial speech from natural speech using LPC coefficients. Spoiler alert, LPC is very bad, but at least we've figured that pole order at around 24-28 is quite optimal for the task

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages