Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

History

Compare Revisions

Revisions

@douglasbagnall douglasbagnall add generated table of phoneme counts for various dictionaries 333527a
@douglasbagnall douglasbagnall espeak-arpabet pages based on commit e7aaa363 (more diphthongs) 1c65ecb
@douglasbagnall douglasbagnall ignore backups~ e9d031b
@douglasbagnall douglasbagnall espeak IPA pronunciations by translated into arpabet (based on e7aaa3638) fda57eb
@douglasbagnall douglasbagnall typo in extracting utterances markup 6ecb4c5
@douglasbagnall douglasbagnall Notes about finding utterances in long featureless recordings 1121d01
@douglasbagnall douglasbagnall Add some more espeak accent test pages b0d6bc8
@douglasbagnall douglasbagnall remove duplicates from espeak example 5f7b1bb
@douglasbagnall douglasbagnall Test dictionary generated by espeak G2P 7572fbd
@douglasbagnall douglasbagnall Word list reports with absent terms noted. Also adding Voxforge dict Voxforge looks a bit like a cmudict fork, with schwa's (ax) added. 4897ee5
@douglasbagnall douglasbagnall Wictionary reads word list (badly, not entirely its fault) 52c54bb
@douglasbagnall douglasbagnall add link to new cmudict wordlist f55dd05
@douglasbagnall douglasbagnall I didn't like that phrase 8b3c0fb
@douglasbagnall douglasbagnall Redo word-lists with proper handling of duplicate entries The previous version got the markdown wrong if the dictionary had more than one definition, as is often the case. New script: \#!/bin/sh WORDSET=$1 DICT=$2 echo "#" looking up $(basename $WORDSET) in $(basename $DICT) echo 'Generated by `' $0 $@ '`'. while read -r line; do echo echo "## $line" echo for word in $line; do grep -i "^$word[([:space:]]" $DICT | while read -r result; do echo " $result" done done done < $WORDSET 80c8908
@douglasbagnall douglasbagnall Add some automatically generated NZE word list readings These were created with the following script -- hopefully they are valid markdown: \#!/bin/sh WORDSET=$1 DICT=$2 echo "# looking up $(basename $WORDSET) in $(basename $DICT)" while read -r line; do echo echo "## $line" echo for word in $line; do echo -n " " grep -i "^$word[([:space:]]" $DICT done done < $WORDSET cb1ccf7
@douglasbagnall douglasbagnall add link to NZE word list af51865
@douglasbagnall douglasbagnall Updated Cmudict's empirical dialect (markdown) 9bd9b6e
@douglasbagnall douglasbagnall See what CMU dict says about the wordlist used in the Canterbury Corpus aeb886e
@douglasbagnall douglasbagnall add some more pairs 01808a8
@douglasbagnall douglasbagnall Add meat-meet in there 5ca8fab
@douglasbagnall douglasbagnall Updated Cmudict's dialect what is it really? (markdown) 1699428
@douglasbagnall douglasbagnall Test cmudict against some minimal and not so minimal pairs and tuples 02598c7
@douglasbagnall douglasbagnall Page linking to various engines 9ed3211
@douglasbagnall douglasbagnall Small update after taking to Bernadette Vine about Wellington Corpus e310193
@douglasbagnall douglasbagnall Updated Acoustic models of New Zealand English (rest) 9c66488
@douglasbagnall douglasbagnall Updated Acoustic models of New Zealand English (rest) ca5b6a8
@douglasbagnall douglasbagnall initial description, adaptation vs creation cc430fb
@douglasbagnall douglasbagnall Initial Commit 7d84537
Something went wrong with that request. Please try again.