search

nassosoassos / sail_align

SailAlign is an open-source software toolkit for robust long speech-text alignment implementing an adaptive, iterative speech recognition and text alignment scheme that allows for the processing of very long (and possibly noisy) audio and is robust to transcription errors. It is mainly written as a perl library but its functionality also depends on freely available software, namely HTK, srilm and sclite.
97 stars 14 forks source link

Network must be specified for recognition from audio - Sampling Rate is 16k #5

Closed saikrishnar closed 9 years ago

saikrishnar commented 9 years ago

Sir,

Can you please suggest on how to deal with the Out of Vocabulary words?

sailalign_oov

nassosoassos commented 9 years ago

I'm not sure how I missed this. So, in the configuration file you have the option to specify a complementary dictionary (check the reference field below, you may add a dictionary there, as shown below): (example from the timit_alignment.cfg) language_modeling => { dictionary => { tool => 'htk', bin_dir => $BIN_DIR, log_file_suffix => 'dlog', reference => [ catfile( $ROOTPATH, 'language', 'cmu_dictionary.dic' ), catfile( $ROOTPATH, 'language', 'timit_dictionary.dic' ), catfile( 'my_path', 'complementary_dictionary.dic' ), ], output_symbols_list => ...

Pronunciations there should be defined in the same format as in the cmu_dictionary.