I plan on learning/using an open source phoneme recognition package called cmusphinx/pocketsphinx.
https://cmusphinx.github.io/wiki/phonemerecognition/
Also there are 39 phonemes (plus a silence) so my design will increase the outputs from 32 to 39.
I have also noticed that these are always sequential so can be represented in 6 bits rather than the previous 32. For practical reasons each character should be represented as 8bit, or 2 hex characters.
Examples of the decoding taken from the cmusphinx is laid out below, with hex values added.
HEX CMUBET IPA Example Translation -- ------ --- ------- ----------- 00 SIL . silence ...word-end [ SIL ] word-start... 01 AA ɑ odd AA D 02 AE æ at AE T 03 AH ʌ hut HH AH T 04 AO ɔ ought AO T 05 AW ɑʊ cow K AW 06 AY ɑɪ hide HH AY D 07 B b be B IY 08 CH ʧ cheese CH IY Z 09 D d dee D IY 0A DH ð thee DH IY 0B EH ɛ Ed EH D 0C ER ɜɹ hurt HH ER T 0D EY eɪ ate EY T 0E F f fee F IY 0F G ɡ green G R IY N 10 HH h he HH IY 11 IH i it IH T 12 IY ɪː eat IY T 13 JH ʤ gee JH IY 14 K k key K IY 15 L l lee L IY 16 M m me M IY 17 N n knee N IY 18 NG ŋ ping P IH NG 19 OW oʊ oat OW T 1A OY ɔɪ toy T OY 1B P p pee P IY 1C R ɹ read R IY D 1D S s sea S IY 1E SH ʃ she SH IY 1F T t tea T IY 20 TH θ theta TH EY T AH 21 UH ʊ hood HH UH D 22 UW u two T UW 23 V v vee V IY 24 W w we W IY 25 Y j yield Y IY L D 26 Z z zee Z IY 27 ZH ʒ seizure S IY ZH ER
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.