Changes between Version 7 and Version 8 of HMMVoiceCreationMary4.0


Ignore:
Timestamp:
09/23/09 16:53:24 (16 years ago)
Author:
marcela_charfuelan
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • HMMVoiceCreationMary4.0

    v7 v8  
    88III) Training of HMM models[[BR]] 
    99IV)  Adding a new HMM voice in the Mary system.[[BR]] 
    10 V)   Creating other voice in German or English (''to train a HMM voice with another speech database'').[[BR]] 
     10V)   Creating other voice in German (''to train a HMM voice with another speech database'').[[BR]] 
    1111VI) (''NEW'') Creating other voice in a language different from German or English (US). 
    1212 
    … …  
    5050- EHMM for automatic labeling, available with festvox-2.1 link: http://festvox.org/download.html [[BR]] 
    5151 
    52 The HTS demo for MARY 4.0 beta, includes a shell script "check_programs.sh" that will help you to check if all the previous  
    53 programs are installed, otherwise it will suggest how they can be installed. 
     52The HTS demo for MARY 4.0 beta, includes a shell script "check_programs.sh" that will help you to check if all the previous programs are installed. 
    5453 
    5554 
    … …  
    6160Where to start? There are three options: 
    6261 
    63 '''a-''' If you would like to try the HTS-SLT-demo (slt voice) for MARY 4.0 from scratch:[[BR]] 
    64 Download the HTS-demo_CMU-ARCTIC-SLT_for_MARY-4.0-beta.tar.gz (92MB) and un-zip, un-tar it in a directory where you will 
    65 train the SLT voice.[[BR]] 
    66  
    67 '''b-''' If you have already created a unit selection voice for MARY, with the SLT data, and want to build a HMM-based voice for that:[[BR]] 
    68 Copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz (112K) in your unit selection voice creation directory and un-zip, un-tar the file.[[BR]]  
    69 If you have already created a unit selection voice for this data, most probably you have already created phonefeatures, phonelab and a mary/features.txt file for that, so you can run steps 1-3 and skip steps 4-11. 
     62'''a-''' If you would like to try the HTS-demo_CMU-ARCTIC-SLT for MARY 4.0 beta from scratch:[[BR]] 
     63Download the HTS-demo_CMU-ARCTIC-SLT_for_MARY-4.0-beta.tar.gz (92MB), unpack the file and go to that directory: 
     64{{{ 
     65   tar -zxvf HTS-demo_CMU-ARCTIC-SLT_for_MARY-4.0-beta.tar.gz 
     66   cd HTS-demo_CMU-ARCTIC-SLT_for_MARY-4.0-beta 
     67}}} 
     68 
     69'''b-''' If you have already created a unit selection voice for MARY, with the CMU-ARCTIC-SLT data, and want to build a HMM-based voice for that,  
     70copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz (112K) in your unit selection voice creation directory and unpack the file: 
     71{{{ 
     72   tar -zxvf HTS-demo_for_MARY-4.0-beta.tar.gz 
     73}}} 
     74If you have already created a unit selection voice for this data, most probably you have already created phonefeatures, phonelab and a mary/features.txt file for that, so you can run steps 1-3, skip steps 4-11 and continue with section III HMM models training. 
    7075 
    7176'''c-''' If you want to create a HMM voice in other language please see the section V or VI below.[[BR]] 
    … …  
    7782./check_programs.sh 
    7883}}}   
    79 This is a simple shell script that will check first, if the programs are available in the PATH and report what is missing. You can provide the paths where you install the programs if they are not found in the PATH. The script will suggest how to install missing programs. 
    80  
     84This is a simple shell script that will check which programs are available in the PATH and report what is missing. You can provide the paths where you have installed the  required programs if they are not found in the PATH. The script will check minimal requirements for programs, versions, suggest how to install missing programs, etc. 
     85If all the necessary programs are installed correctly you can continue with step 2.  
    8186 
    8287'''2-''' Run the VoiceImport program 
    8388 
    84 First of all you need to set your MARY_BASE directory and then run the program:  [[BR]] 
     89If you have installed MARY 4.0 beta and the voicebuilding component, you can start the VoiceImport program from '''Applications -> OpenMary -> Voice import tools'''[[BR]] 
     90Otherwise you can run it on a terminal in your working directory (the directory where you have unpacked the HTS demo for MARY 4.0 beta), first you need to set your MARY_BASE directory and then run the voiceimport.jar program: 
    8591{{{ 
    8692   export MARY_BASE="/dir/to/openmary" 
    … …  
    8894}}} 
    8995 
    90 When starting the voiceimport tools please provide information for: 
     96When starting the voiceimport tools, go to your working directory (the directory where you have unpacked the HTS demo for MARY 4.0 beta) and provide information for: 
    9197{{{ 
    9298  db.gender    = female 
    9399  db.locale    = en_US 
    94100  db.marybase  = /path-to-MARY_BASE/ or /path-to-openmary/ 
    95   db.voicename = hsmm-slt 
     101  db.voicename = slt-hsmm 
    96102}}} 
    97103If you are not familiar or have problems with the VoiceImport program, please read the instructions in the Voice Import Tools 
    … …  
    106112}}} 
    107113 
    108 '''3-''' Run the HMMVoiceDataPreparation of the HMM Voice Trainer group to check if text, wav and data/raw files are available and in the correct paths. 
    109 If just data/raw is provided, the program will do the conversion.  If no text files are available but data/utts in festival format, the program will do the conversion as well. 
     114'''3-''' Run the HMMVoiceDataPreparation of the HMM Voice Trainer group to check if text, wav or raw files are available and in the correct paths. 
     115If just raw is provided, the program will do the conversion.  If no text files are available but utts in festival format, the program will do the conversion as well. 
    110116 
    111117'''4-''' Run the AllophonesExtractor of the Automatic Labeling group to create the '''prompt_allophones''' directory required in the next step. This component requires the MARY server. [[BR]] 
    … …  
    243249 
    244250''' 
    245 === V) Creating other voice in German or English. === 
     251=== V) Creating other voice in German === 
    246252''' 
    247253 
    … …  
    250256  * a wav or raw directory with the speech files you will use for training the German voice. [[BR]] 
    251257  * transcriptions of the files, one text file per speech file, or transcriptions in festival format if available. [[BR]] 
    252 then copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz file in the directory where you have your wav and transcription data and un-zip, un-tar the file.[[BR]] 
     258then copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz file in the directory where you have your wav and transcription data and unpack the file: 
     259{{{ 
     260   tar -zxvf HTS-demo_for_MARY-4.0-beta.tar.gz 
     261}}} 
    253262 
    254263Once you have unpacked the HTS demo for MARY 4.0 beta, follow the instructions as normal from step 1. Provide general settings for: 
    … …  
    261270 
    262271 
    263 - If you have already created a German unit selection voice for MARY and want to build a HMM-based voice for that:[[BR]] 
    264 copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz (112K) in your unit selection voice creation directory and un-zip, un-tar the file.[[BR]]  
    265  
    266 If you have already created a unit selection voice for German, most probably you have already created phonefeatures, phonelab and a mary/features.txt file for that, so you can run steps 1-3 and skip steps 4-11. 
     272- If you have already created a German unit selection voice for MARY and want to build a HMM-based voice for that, copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz (112K) in your unit selection voice creation directory and unpack the file: 
     273{{{ 
     274   tar -zxvf HTS-demo_for_MARY-4.0-beta.tar.gz 
     275}}} 
     276 
     277If you have already created a unit selection voice for German, most probably you have already created phonefeatures, phonelab and a mary/features.txt file for that, so you can run steps 1-3, skip steps 4-11 and continue with section III HMM models training. 
    267278 
    268279 
    … …  
    277288- '''Phoneme set''':  contained in MARY_BASE/lib/modules/xx/lexicon/allophones.xx.xml , where xx corresponds to the new language. 
    278289 
    279 - After creating the minimal components, follow the steps in section V.  
     290- After creating the minimal components, you will need wav files (in a wav directory) and the corresponding transcriptions (one file per wav file in a text directory). [[BR]] 
     291Then copy the openmary/lib/hts/HTS-demo_for_MARY-4.0-beta.tar.gz file in the directory where you have your wav and transcription data and unpack the file: 
     292{{{ 
     293   tar -zxvf HTS-demo_for_MARY-4.0-beta.tar.gz  
     294}}} 
     295 
     296Once you have unpacked the HTS demo for MARY 4.0 beta, follow the instructions as normal from step 1. Provide general settings for: 
     297{{{ 
     298   db.gender    =  male  (or female) 
     299   db.locale    =  new_language locale (according to your minimal NLP components, ex. tr for Turkish, te for Telugu, etc.) 
     300   db.marybase  =  /path/to/mary/base/ 
     301   db.voicename =  new_language_voice_name 
     302}}} 
     303 
     304 
     305   
    280306 
    281307[[BR]]