hello list!
for an upcoming project i will need a speech recognition system (trained to one speaker with a limited set of words) which makes the data available to pd (best would be a OSC stream of the recognized words). i've found sphinx, a opensource recognition system and was thinking about adapting it, but my coding abilities aren't sufficient ;) and i have major problems getting the system alone to work... has anyone tried something similar yet? or can someone point me to possible solutions? maybe a simple commandline tool, and a little something for pd that reads the output?
any help appreciated! best, leo
depart.at
for an upcoming project i will need a speech recognition system (trained to one speaker with a limited set of words) which makes the data available to pd (best would be a OSC stream of the recognized words). can someone point me to possible solutions?
-Perhaps this could be the solution:
"ComParser is computer software for the recognition of audio in realtime. The software can be trained to pseudo-scorefollow (certain) musical compositions. The word pseudo-scorefollowing means to say ComParser actually uses no symbolic score at all, and, it only vagely knows the exact location during following. http://kmt.hku.nl/~pieter/SOFT/CMP/doc/cmp.html
The software is written in ANSI/ISO C as far as possible. Sourcecode, binaries, documentation and examples may be downloaded from http://kmt.hku.nl/~pieter/SOFT/CMP/.
ComParser is free open source software developed by the Schreck Ensemble. It has been presented during a masterclass in computer music at the Centro Ricerche Musicali in Rome, june 21, 2003; and applied in performances of Hans van Eck's compostion Poemi Eterni, Tiziana Pintus playing the violin (in Rome and Sassari in 2003, and in De Witte Dame in Eindhoven in 2004)".
There is a ComParser-external for PD for x86 processors: http://kmt.hku.nl/~pieter/SOFT/CMP/src/pd/comparser~.pd_linux-gcc comparser~.pd_linux-gcc 291 400 bytes Pure Data external, version 1.41, to run under Linux on x86 processors, Compiled with GNU GCC. Remove the trailing '-gcc'.
Also a ComParser-external for PD for PPC processors: http://kmt.hku.nl/~pieter/SOFT/CMP/src/pd/comparser~.pd_darwin-gcc comparser~.pd_darwin-gcc 295 484 bytes Pure Data external, version 1.39, to run under Mac OSX. Compiled with GNU GCC. Remove the trailing '-gcc'.
you will have to add OSC to it yourself in PD.
good luck,
AvS
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .....................................................................
` |Schreck Ensemble . . . . . . . . . . . . . . . . . . . . +
` |# -laboratory for live electro-acoustic music- # |
| http://www.schreck.nl/ |
| http://www.xs4all.nl/~schreck/ |
` *===========================================================++
` |Compositions http://www.xs4all.nl/~schreck/html/compo.html |
` |Samples http://www.xs4all.nl/~schreck/html/samp.html |
` |Patches http://www.xs4all.nl/~schreck/html/pat.html |
` |Videos http://www.xs4all.nl/~schreck/html/video.html |
` |Scores http://www.xs4all.nl/~schreck/html/scores.html |
*===========================================================++
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .....................................................................
words). i've found sphinx, a opensource recognition system and was thinking about adapting it, but my coding abilities aren't sufficient ;) and i have major problems getting the system alone to work... has anyone tried something similar yet? or can someone point me to possible solutions?
well, i was thinking on a sphinx external for pd, since i also need it for a project, i'm working on ... still, it's very low priority on my todo list ... but maybe we can try to combine our efforts (better this way than reinventing the wheel twice) ...
cheers ... tim