Skip to content

dltaixlt/AlignmentDuration

 
 

Repository files navigation

Copyright © 2015,2016  Music Technology Group - Universitat Pompeu Fabra

NAME: AlignmentDuration

DESCRIPTION 
 tool for Aligning lyrics to audio automatically using Hidden Markov Models with Explicit Duration

LICENSE:
AlignmentDuration is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation (FSF), either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program.  If not, see http://www.gnu.org/licenses/



BUILD INSTRUCTIONS: 
git clone -b IntegratedPycompMusic https://github.com/georgid/AlignmentDuration.git
python setup.py install


* htkModelParser
git clone https://github.com/georgid/htkModelParser.git
python setup.py install


* sci-kit-learn (branch with fixed underflow issues)
git clone https://github.com/georgid/scikit-learn 
python setup.py install

* utilsLyrics
git clone https://github.com/georgid/utilsLyrics.git
python setup.py install


-----
* Evaluation (optional if model evaluation step needed, see Doit.py )
git clone https://github.com/georgid/AlignmentEvaluation.git

* HMM (optional if approach with no-duration modeling needed)
https://github.com/MTG/HMM.git


-------------------------------------------------------


USAGE :
pycompmusic/compmusic/extractors/makam/lyricsalign.py

to test without pycommusic: 
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/test/testLyricsAlign.py


all from here is DEPRACATED: 
EXAMPLE: doitOneChunk()
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitOneChunk.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/nihavent--sarki--duyek--bir_ihtimal--osman_nihat_akin/ /Users/joro/Documents/Phd/UPF/adaptation_data_soloVoice/ISTANBUL/idil/05_Semahat_Ozdenses_-_Bir_Ihtimal_Daha_Var_0_zemin_from_69_5205_to_84_2 True False 0.97 False 2

# with re-synthesized audio
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitOneChunk.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/ussak--sarki--duyek--aksam_oldu_huzunlendim--semahat_ozdenses/  /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/ussak--sarki--duyek--aksam_oldu_huzunlendim--semahat_ozdenses/06_Semahat_Ozdenses_-_Aksam_Oldu_Huzunlendim/06_Semahat_Ozdenses_-_Aksam_Oldu_Huzunlendim_1_zemin_from_62_393939_to_77_827796 False 0.97 False 2

python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitOneChunk.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/muhayyerkurdi--sarki--duyek--ruzgar_soyluyor--sekip_ayhan_ozisik_short/  /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/muhayyerkurdi--sarki--duyek--ruzgar_soyluyor--sekip_ayhan_ozisik_short/1-05_Ruzgar_Soyluyor_Simdi_O_Yerlerde/1-05_Ruzgar_Soyluyor_Simdi_O_Yerlerde_2_zemin_from_16_459_to_38_756510 True 0.97 False 2

EXAMPLE: doitOneRecording() 
# with re-synthesized audio
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitOneRecording.py  /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/ussak--sarki--duyek--aksam_oldu_huzunlendim--semahat_ozdenses/ /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis//ussak--sarki--duyek--aksam_oldu_huzunlendim--semahat_ozdenses/06_Semahat_Ozdenses_-_Aksam_Oldu_Huzunlendim/ 06_Semahat_Ozdenses_-_Aksam_Oldu_Huzunlendim_ True True 0.97 False 2 True

EXAMPLE: doitAllRecordings()
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitAllRecordings.py /Users/joro/Documents/Phd/UPF/  False False 0.97 False 2
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitAllRecordings.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/   /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/ False True 0.97 False 2

# with synthesized audio
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitAllRecordings.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/   /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data-synthesis/ False True 0.97 False 2

(on hpc)
see runAllRecordings.sh


EXAMPLE: doitAllRecParams()
python /Users/joro/Documents/Phd/UPF/voxforge/myScripts/AlignmentDuration/doitAllRecParams.py /Users/joro/Documents/Phd/UPF/turkish-makam-lyrics-2-audio-test-data/  /Users/joro/Documents/Phd/UPF/adaptation_data_soloVoice/ISTANBUL/

(on hpc)
python AlignmentDuration/doitAllRecParams.py /homedtic/georgid/turkish-makam-lyrics-2-audio-test-data/  /homedtic/georgid/ISTANBUL/


loading persistent files: 
extension to load observation probabilities for all states and all observations for a given file is in 
Decoder.Decoder.decodeAudio

About

lyrics-to-audio-alignement system. Decoding with Viterbi forced alignment. Note duration aware decoding

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 87.5%
  • Jupyter Notebook 9.2%
  • C 2.5%
  • Shell 0.8%