ARABIC SPEECH RECOGNITION USING SPHINX

[hasan ali gamal al-kaf]

Hi Nickolay
when i want to convert my audio to text using pocketsphinx

pocketsphinx_continuous -infile
'/home/hasan/Desktop/final1/sphinxtrain/an4/wav/an4_clstk/aabbccp/an1-aabbccp-b.wav'
-jsgf '/home/hasan/Desktop/final1/sphinxtrain/etc/an4.jsgf' -dict
'/home/hasan/Desktop/final1/sphinxtrain/etc/an4.dic'

but i got this error
ERROR: "dict.c", line 195: Line 1: Phone 'I' is mising in the acoustic
model; word 'BISMI' ignored
ERROR: "dict.c", line 195: Line 2: Phone 'A' is mising in the acoustic
model; word 'LAHI' ignored
ERROR: "dict.c", line 195: Line 3: Phone 'HI' is mising in the acoustic
model; word 'RAHMANI' ignored
ERROR: "dict.c", line 195: Line 4: Phone 'HI' is mising in the acoustic
model; word 'RAHIM' ignored
ERROR: "dict.c", line 195: Line 5: Phone 'HI' is mising in the acoustic
model; word 'RAHIM(2)' ignored
ERROR: "dict.c", line 195: Line 6: Phone 'HI' is mising in the acoustic
model; word 'RAHIM(3)' ignored
ERROR: "dict.c", line 195: Line 7: Phone 'KO' is mising in the acoustic
model; word 'KOUL' ignored
ERROR: "dict.c", line 195: Line 8: Phone 'HI' is mising in the acoustic
model; word 'HOUWA' ignored
ERROR: "dict.c", line 195: Line 9: Phone 'HI' is mising in the acoustic
model; word 'LLAHOU' ignored
ERROR: "dict.c", line 195: Line 10: Phone 'H' is mising in the acoustic
model; word 'AHAD' ignored
ERROR: "dict.c", line 195: Line 11: Phone 'H' is mising in the acoustic
model; word 'AHAD(2)' ignored
ERROR: "dict.c", line 195: Line 12: Phone 'HI' is mising in the acoustic
model; word 'ALLAHOU' ignored
ERROR: "dict.c", line 195: Line 13: Phone 'SO' is mising in the acoustic
model; word 'SAMAD' ignored
ERROR: "dict.c", line 195: Line 14: Phone 'A' is mising in the acoustic
model; word 'YALED' ignored
ERROR: "dict.c", line 195: Line 15: Phone 'A' is mising in the acoustic
model; word 'WA' ignored
ERROR: "dict.c", line 195: Line 16: Phone 'A' is mising in the acoustic
model; word 'LAM' ignored
ERROR: "dict.c", line 195: Line 17: Phone 'OU' is mising in the acoustic
model; word 'YOULAD' ignored
ERROR: "dict.c", line 195: Line 18: Phone 'A' is mising in the acoustic
model; word 'YAKOUN' ignored
ERROR: "dict.c", line 195: Line 19: Phone 'A' is mising in the acoustic
model; word 'LAHOU' ignored
ERROR: "dict.c", line 195: Line 20: Phone 'OU' is mising in the acoustic
model; word 'KOUFOUAN' ignored
INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones
INFO: dict.c(336): 0 words read
INFO: dict.c(358): Reading filler dictionary:
/usr/local/share/pocketsphinx/model/en-us/en-us/noisedict
INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones
INFO: dict.c(361): 5 words read
INFO: dict2pid.c(396): Building PID tables for dictionary
INFO: dict2pid.c(406): Allocating 42^3 * 2 bytes (144 KiB) for word-initial
triphones
INFO: dict2pid.c(132): Allocated 42672 bytes (41 KiB) for word-final
triphones
INFO: dict2pid.c(196): Allocated 42672 bytes (41 KiB) for single-phone word
triphones
INFO: jsgf.c(706): Defined rule: <an4.g00000>
INFO: jsgf.c(706): Defined rule: PUBLIC <an4.an4>
INFO: fsg_model.c(207): Computing transitive closure for null transitions
INFO: fsg_model.c(269): 0 null transitions added
INFO: fsg_search.c(227): FSG(beam: -1080, pbeam: -1080, wbeam: -634; wip:
-26, pip: 0)
ERROR: "fsg_search.c", line 141: The word 'WA' is missing in the dictionary</an4.an4></an4.g00000>

the the word WA is in the dictionary

ERROR: "dict.c", line 195: Line 1: Phone 'I' is mising in the acoustic
model; word 'BISMI' ignored
ERROR: "fsg_search.c", line 141: The word 'WA' is missing in the dictionary

could you please help me . i am in the last step in my project

i will attach my dictionay , cfg ,jsgf ,align and cfg

thank you so much

[Nickolay V. Shmyrev]

You need to specify the acoustic model with -hmm option.

[hasan ali gamal al-kaf]

HI Nickolay

When i want to use gui interface for my pocketsphinx

For example, first of all i want to give the information of the persons who
wants to use the software
then if he wants batch mode or real time
if he chooses batch mode upload the audio and he gets the text and the
accuracy

Can i make this idea using gui interface visual basic on ubuntu ? is there
any better qui ?

Thank you so much

[Nickolay V. Shmyrev]

You can better create GUI with Python, it is much better than Visual Basic, see tutorials here:

http://www.pygtk.org/pygtk2tutorial/ch-GettingStarted.html#sec-HelloWorld

[hasan ali gamal al-kaf]

Hi Nickolay ,

I have idea i want to apply it which I want my system to to follow what the
reader say then if the reader stop .it remember him what next text is or
next word ?

Can i apply this idea using cmusphinx ? if not which software can i use to
make this idea?

Thank you so much

[Nickolay V. Shmyrev]

Hello Hasan

You can make such system with CMUSphinx, you only need to recognize with text-biased langauge model and track the last pronounced word. That should work reliably well if the text is small.

[hasan ali gamal al-kaf]

So can he remember him by voice. so that mean after track the last
pronounced word . i want the system read the word that forget so he change
from text to sound (only the word that forget ) and how fast will read the
word ? how long i will take to achive this idea after i already change from
sound to text using pocketsphinx

what if i have long text ? is there any software

thank you so much

[Nickolay V. Shmyrev]

I'm sorry, it is hard to understand what you ask.

If you mean you want to make a hint to the user who reads the text from memory, you usually wait for some time and then synthesize the word. It should not take long, just a second of delay and you make a hint.

It is not hard to build such system for experienced developer, I have no idea what are your skills so can not tell for you.

If text is long you usually can split it on chapter and process each chapter separately. Once chapter of 15-20 minutes speech can be recognized accurately.

[hasan ali gamal al-kaf]

Thank you Nickolay

For example, i have poem and I want the blind kids to memories

So when the blind kids read the poem i want my system follow them . then
if the blind kids forget what is the next word in the poem .my system tell
him by voice what is the next word

what is the steps to make this system ?

Thank you so much Nickolay

[Nickolay V. Shmyrev]

what is the steps to make this system ?

1. Open a code editor
2. Start writing code
3. Write code
4. Debug mistakes
5. Give system to the users

[hasan ali gamal al-kaf]

HI Nickolay

could you explain to me which code editor is easy for me to apply my idea
and is there any tutorial should i start learning ?
because i am new in a code editor .

my idea

For example, i have poem and I want the blind kids to memories

So when the blind kids read the poem i want my system follow them . then
if the blind kids forget what is the next word in the poem .my system tell
him by voice what is the next word

[Nickolay V. Shmyrev]

Hi Hasan

Best IDE to choose depends on the programming language and platform of your preference.

[hasan ali gamal al-kaf]

HI Nickolay

i will choose android speech recognition for ide

the flowchart of the code
1- code for sound to text using pocketsphinx
2- code for delay
3- code for text to sound

is it my flow chart correct ? or there are another methods should i
understand  to do the idea?


the idea

1 i have poem and I want the blind kids to memories

So when the blind kids read the poem i want my system follow them . then
if the blind kids forget what is the next word in the poem .my system tell
him by voice what is the next word

thank you so much



[Nickolay V. Shmyrev]

is it my flow chart correct ? or there are another methods should i

I think the flow chart is correct