speech1 - voxforge.org

Speech Recognition Engines

Flat

speech1

User: pvs
Date: 6/8/2009 8:44 pm

Views: 5056
Rating: 3

plz help me any one.

I intrested to do speaker adaptation with an4 and rm1 database.

I find one pdf file

http://www.2shared.com/file/6155477/aa760d42/doh_thesis.html

I understand the equations of PC-MLLR,but where i can changes are do in Sphinx3 exactly are not proprely understand.plz any one help to me.

In sphinx3 have map_adapt folder is their.

in some doc is given mllr_adapt.I think to create mllr_adapt floder is my work.plz any one help to me.

Enhancements to Transformation-Based Speaker Adaptation:

Principal Component MLLR.

--- (Edited on 6/8/2009 8:44 pm [GMT-0500] by Visitor) ---

Re: speech1

User: nsh
Date: 6/9/2009 12:13 am

Views: 65
Rating: 3

mllr_adapt is a part of sphinxtrain. See

http://www.speech.cs.cmu.edu/cmusphinx/moinmoin/AcousticModelAdaptation

--- (Edited on 6/9/2009 12:13 am [GMT-0500] by nsh) ---

Re: speech1

User: pvs
Date: 6/9/2009 1:11 am

Views: 63
Rating: 2

hi nsh

Thanks to given a reply to me,but you don't leave in middle because i want full information about this plz help to me.Because No one is help to me.

I want to this as my Mtech project.

I seen map_adapt folder in sphinxtrain.My project related doc also given to u.

here I want is PC-MLLR method.where the changes are done .

Atleast tell if it is possible or not.

Thanks nsh.plz give a reply whatever it is.

Thanks somuch.

--- (Edited on 6/9/2009 1:11 am [GMT-0500] by Visitor) ---

Re: speech1

User: nsh
Date: 6/9/2009 1:26 am

Views: 55
Rating: 3

If you want to get answers, ask better questions, I don't understand what are you asking about. Spend 30 minutes to formulate the question and you'll get the anwer.

--- (Edited on 6/9/2009 1:26 am [GMT-0500] by nsh) ---

Re: speech1

User: pvs
Date: 6/9/2009 4:13 am

Views: 1880
Rating: 3

hi nsh

sorry nsh.

I taken an4 and rm1 databases and done the normal recognition process by using sphinx tutorial.

Now i want do speaker adaptation

In that speaker adaptation I want to implement

Principal Components Analysis MLLR method.That doc link also

attached last mail.

My question is where I implement PC-MLLR method in SphinxTrain.

sorry nsh.I asking question with lack knowledge try to understand.

Thanks .plz give reply.

--- (Edited on 6/9/2009 4:13 am [GMT-0500] by Visitor) ---

Previous • Next •


Username	Password