Okay this may be a stupid question...
I'm trying to get the pocketsphinx working for my iPhone.. so far "I believe"
I manage to compile both base & pocketsphinx for iPhone... Now I need some
help with coding... Can any one give me some sample in Objective C! I'm kind
of lost...
Thank in advance for any help....
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hm... nothing so far... okay!
Can any one help me with why this error occurs?
"_iconv_close", referenced from:
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
"_iconv", referenced from:
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
"_iconv_open", referenced from:
_ngram_model_recode in libsphinxbase.a(ngram_model.o)
ld: symbol(s) not found
collect2: ld returned 1 exit status
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
nshmyrev, thanks for the reply... I didn't know it would work without ICONV!
anyway... I couldn't find the config.h, instead I disabled the "AM_ICONV" in
config.in and compiled all over again!..
Hey rjeevan, just to tell you that I went exactly through the same issues as
you, and I'm currently at the same state. (For Iconv, I moved the file
iconv.m4 in the right directory).
What I'm gone do from now on is to rewrite the equivalent of the
pocketsphinxmigration example in objective C using the audio queue services
from Apple. I think (and hope) I can manage to call the sphinx's speechToText
routine (ps_get_hyp as far as I understood) from within my program. Will let
you know.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Okay finally, I finished my Project with PocketSphinx and all went well... I'm
using decode_raw and passing recorded audio file to decode and it's working
fine in iPhone 3GS... It's slow and taking time to recognise but WORKING!....
I will see if I can write about How to get it working in Iphone later on...
that may help other who's interested too....
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Rjeevan Im interested in this project as well and im thinking ways to
improve it . What you have found is to record a file and pass that file to be
decoded. I found what makes it taking a long time because the iphone has to
write the buffer onto a file on the hard disk . So im wondering if we can make
the audio queue write to the ram instead of the hard disk? im researching
about this right now. Will let you know what i will find. Hope we can share
info...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
arckit, thanks... My time was limited... this was my final year project and I
started 2 weeks before the due date... I;m totally new to XCODE (Objective C)
and on C my knowladge is very basic! I tried to use the Audio Queue get buffer
and pass it on to ps, but it didn;t work for me and no time for further
research... so I did this way to make my life easier!.... :)
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Okay, I've written something about how to get PocketSphinx working inside
iPhone using pre-recorded audio and added my xcode project to be downloaded…
My english is not in professional level, but I'm sure you can manage.
Has anyone figured out how to get this to compile for xcode 3.2.3?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2010-06-21
If you open up build properties and set 'Architectures' to "armv6" instead of
"armv6 armv7" it will build successfully. There's a bug in XCode 3.2.3 that's
aggravating this issue, and if you just build for armv6 the issue will go
away. It will not be optimized for iPhone 4, but it will work. I'm waiting to
submit an official fix to see what apple does with it.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Okay this may be a stupid question...
I'm trying to get the pocketsphinx working for my iPhone.. so far "I believe"
I manage to compile both base & pocketsphinx for iPhone... Now I need some
help with coding... Can any one give me some sample in Objective C! I'm kind
of lost...
Thank in advance for any help....
hm... nothing so far... okay!
Can any one help me with why this error occurs?
This error occurs because you don't link to iconv library. You could undef
HAVE_ICONV in config.h file in order to compile without iconv.
nshmyrev, thanks for the reply... I didn't know it would work without ICONV!
anyway... I couldn't find the config.h, instead I disabled the "AM_ICONV" in
config.in and compiled all over again!..
Now I need to figure out how connect the Microphone instead of raw wave file
as suggested in http://cmusphinx.sourceforge.net/wiki/pocketsphinxmigration
do you know of nay tutorials for streaming?
Hey rjeevan, just to tell you that I went exactly through the same issues as
you, and I'm currently at the same state. (For Iconv, I moved the file
iconv.m4 in the right directory).
What I'm gone do from now on is to rewrite the equivalent of the
pocketsphinxmigration example in objective C using the audio queue services
from Apple. I think (and hope) I can manage to call the sphinx's speechToText
routine (ps_get_hyp as far as I understood) from within my program. Will let
you know.
heyfloyd, thanks for the update... please do share if you find something...
will do the same!...
Okay finally, I finished my Project with PocketSphinx and all went well... I'm
using decode_raw and passing recorded audio file to decode and it's working
fine in iPhone 3GS... It's slow and taking time to recognise but WORKING!....
I will see if I can write about How to get it working in Iphone later on...
that may help other who's interested too....
Hi Rjeevan Im interested in this project as well and im thinking ways to
improve it . What you have found is to record a file and pass that file to be
decoded. I found what makes it taking a long time because the iphone has to
write the buffer onto a file on the hard disk . So im wondering if we can make
the audio queue write to the ram instead of the hard disk? im researching
about this right now. Will let you know what i will find. Hope we can share
info...
arckit, thanks... My time was limited... this was my final year project and I
started 2 weeks before the due date... I;m totally new to XCODE (Objective C)
and on C my knowladge is very basic! I tried to use the Audio Queue get buffer
and pass it on to ps, but it didn;t work for me and no time for further
research... so I did this way to make my life easier!.... :)
Okay, I've written something about how to get PocketSphinx working inside
iPhone using pre-recorded audio and added my xcode project to be downloaded…
My english is not in professional level, but I'm sure you can manage.
download it at: http://www.rajeevan.co.uk/pocketsphinx_in_iphone/
Has anyone figured out how to get this to compile for xcode 3.2.3?
If you open up build properties and set 'Architectures' to "armv6" instead of
"armv6 armv7" it will build successfully. There's a bug in XCode 3.2.3 that's
aggravating this issue, and if you just build for armv6 the issue will go
away. It will not be optimized for iPhone 4, but it will work. I'm waiting to
submit an official fix to see what apple does with it.