I've installed pocket sphinx on android and tested the pochetsphinxdemo with
the en_US model but it's not recognizing well perhaps because it's only
concidering the american accent in prononciation ??
2)
Now i want the app to recognize french because tha's what i need in my
project.
I downloaded the french model , created the fr directory in both hmm and lm in
the sdcard of the phone and pushed to them what i downloaded. I've also
updated in the RecognizerTask.java the following paths
But now when i launch the app on the phone it's taking too long to start and
after pushing the button " hold and speak" and speaking it is only the
progress circle " recognizing speech" appearing and it still like that.
Have i missed other chages in the code or are there another thing to do.
Thank for helping.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
thanks for answering
so it's impossible to work with this model even with a galaxy S3 ?
and is it possible to modify the french3g62K.lm.dmp instead of creating a new
model ?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm in the same situation.
I followed the tutorial to compile pocketSphinx on android. I have bad result
( every word is wrong ... ) , I tried too simplify my model using lmttool htt
p://www.speech.cs.cmu.edu/tools/lmtool.html ( I have a dictionnary composed of 5 word )
and it always give me the wrong answer.
I don't know where the problem is , but i'll keep digging , if I find
something I will post it here.
ps : with the english model and my new dictionnary i'm at a RT of 1.2
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hi,
i tried to create a samaller lm by following the tutorial
but the application is still taking so much time to start and now when i push
the button to talk the app ic closed .
i've just create the file .lm.dmp and push it to the folder containing the old
lm and i updated the line
i thiks it is not necessary to make a new dictionnaire no ?
Do you have any idea why when using lm and dict downloaded it's not
recognizing french and english speech ?? is it because i'm working with a
mobile device ?
thank a lot for helping
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
it still take a long time to launch and with the French model i created no
word is recognized it never shows a word after speaking !
Try to run the same process on desktop with pocketsphinx_continuous, check the
times and accuracy, then move to Android.
To get more detailed answer provide more information - what exact data files
are you using, which file are you trying to recognize. How long exactly does
recognition take and what is displayed in that case. What is the output in
pocketsphinx.log on device or what is the output of the pocketsphinx tool on
desktop. The more information you provide the faster you get the anwer
Please avoid posting duplicate messages on this forum with the same question.
It's not effective to spam this forum, it even reduces your chances to get the
answer.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I've seen this question in FAQ but that mean in my case that there is a
specifique configuration to do for the decoder because my accuracy is very low
. In this case how can i do this please ? is there any tuto for this ?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
" The first thing you need to understand if your accuracy just lower than
expected or very low. If it's very low most likely you misconfigured the
decoder. "
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi
1)
I've installed pocket sphinx on android and tested the pochetsphinxdemo with
the en_US model but it's not recognizing well perhaps because it's only
concidering the american accent in prononciation ??
2)
Now i want the app to recognize french because tha's what i need in my
project.
I downloaded the french model , created the fr directory in both hmm and lm in
the sdcard of the phone and pushed to them what i downloaded. I've also
updated in the RecognizerTask.java the following paths
c.setString("-hmm",
"/mnt/sdcard/edu.cmu.pocketsphinx/hmm/fr/lium_french_f2");
c.setString("-dict",
"/mnt/sdcard/edu.cmu.pocketsphinx/lm/fr/frenchWords62K.dic");
c.setString("-lm",
"/mnt/sdcard/edu.cmu.pocketsphinx/lm/fr/french3g62K.lm.dmp");
But now when i launch the app on the phone it's taking too long to start and
after pushing the button " hold and speak" and speaking it is only the
progress circle " recognizing speech" appearing and it still like that.
Have i missed other chages in the code or are there another thing to do.
Thank for helping.
French language model french3g62K.lm.dmp is too big for the mobile device.
Create smaller language model from the texts you need or use grammar.
thanks for answering
so it's impossible to work with this model even with a galaxy S3 ?
and is it possible to modify the french3g62K.lm.dmp instead of creating a new
model ?
doing adaptation of the french acoustic model would resolve my problem ?
It's not impossible it's slow
It's very easy to create a new model, see the tutorial
http://cmusphinx.sourceforge.net/wiki/tutoriallm
It's not relevant here, the issue is a big language model which is
specifically designed for broadcast news transcription, not the acoustic model
ok, thank a lot for your helpful reply
Hi
I'm in the same situation.
I followed the tutorial to compile pocketSphinx on android. I have bad result
( every word is wrong ... ) , I tried too simplify my model using lmttool htt
p://www.speech.cs.cmu.edu/tools/lmtool.html ( I have a dictionnary composed of 5 word )
and it always give me the wrong answer.
I don't know where the problem is , but i'll keep digging , if I find
something I will post it here.
ps : with the english model and my new dictionnary i'm at a RT of 1.2
perhaps there is a kind of a configuration to do ? even with the french dic
there is no word well recognized !!
hi,
i tried to create a samaller lm by following the tutorial
but the application is still taking so much time to start and now when i push
the button to talk the app ic closed .
i've just create the file .lm.dmp and push it to the folder containing the old
lm and i updated the line
"/mnt/sdcard/edu.cmu.pocketsphinx/lm/fr/newfile.lm.DMP");
i thiks it is not necessary to make a new dictionnaire no ?
Do you have any idea why when using lm and dict downloaded it's not
recognizing french and english speech ?? is it because i'm working with a
mobile device ?
thank a lot for helping
You can check the log file pocketsphinx.log for details. Maybe you did
something wrong, there will be an error in the log
Alternatively, you can try to run the same model on desktop first
You do not need to change the dictionary.. You can cut the words in the
dictionary though to make it faster to load.
it still take a long time to launch and with the French model i created no
word is recognized it never shows a word after speaking !
Try to run the same process on desktop with pocketsphinx_continuous, check the
times and accuracy, then move to Android.
To get more detailed answer provide more information - what exact data files
are you using, which file are you trying to recognize. How long exactly does
recognition take and what is displayed in that case. What is the output in
pocketsphinx.log on device or what is the output of the pocketsphinx tool on
desktop. The more information you provide the faster you get the anwer
Please avoid posting duplicate messages on this forum with the same question.
It's not effective to spam this forum, it even reduces your chances to get the
answer.
Sorry for duplicating the message but this post contains 2 problems that's why
i posted 2 different posts .
I'm not trying to recognize a file in the android app i push the button and
speak and it should recognize what i said i'm not using a file
for example when using english i use :
acoustic model : hub4wsj_sc_8k
dictionary : hub4.5000.dic
language model : hub4.5000.DMP
i run the app on the phone and i push the button to talk i said : hello how
are you
it displays : no no ( wich is very different from what i said )
the log file :
INFO: cmd_ln.c(691): Parsing command line:
Current configuration:
-agc none none
-agcthresh 2.0 2.000000e+00
-alpha 0.97 9.700000e-01
-ascale 20.0 2.000000e+01
-aw 1 1
-backtrace no no
-beam 1e-48 1.000000e-48
-bestpath yes yes
-bestpathlw 9.5 9.500000e+00
-bghist no no
-ceplen 13 13
-cmn current current
-cmninit 8.0 8.0
-compallsen no no
-debug 0
-dict
-dictcase no no
-dither no no
-doublebw no no
-ds 1 1
-fdict
-feat 1s_c_d_dd 1s_c_d_dd
-featparams
-fillprob 1e-8 1.000000e-08
-frate 100 100
-fsg
-fsgusealtpron yes yes
-fsgusefiller yes yes
-fwdflat yes yes
-fwdflatbeam 1e-64 1.000000e-64
-fwdflatefwid 4 4
-fwdflatlw 8.5 8.500000e+00
-fwdflatsfwin 25 25
-fwdflatwbeam 7e-29 7.000000e-29
-fwdtree yes yes
-hmm
-input_endian little little
-jsgf
-kdmaxbbi -1 -1
-kdmaxdepth 0 0
-kdtree
-latsize 5000 5000
-lda
-ldadim 0 0
-lextreedump 0 0
-lifter 0 0
-lm
-lmctl
-lmname default default
-logbase 1.0001 1.000100e+00
-logfn
-logspec no no
-lowerf 133.33334 1.333333e+02
-lpbeam 1e-40 1.000000e-40
-lponlybeam 7e-29 7.000000e-29
-lw 6.5 6.500000e+00
-maxhmmpf -1 -1
-maxnewoov 20 20
-maxwpf -1 -1
-mdef
-mean
-mfclogdir
-min_endfr 0 0
-mixw
-mixwfloor 0.0000001 1.000000e-07
-mllr
-mmap yes yes
-ncep 13 13
-nfft 512 512
-nfilt 40 40
-nwpen 1.0 1.000000e+00
-pbeam 1e-48 1.000000e-48
-pip 1.0 1.000000e+00
-pl_beam 1e-10 1.000000e-10
-pl_pbeam 1e-5 1.000000e-05
-pl_window 0 0
-rawlogdir
-remove_dc no no
-round_filters yes yes
-samprate 16000 1.600000e+04
-seed -1 -1
-sendump
-senlogdir
-senmgau
-silprob 0.005 5.000000e-03
-smoothspec no no
-svspec
-tmat
-tmatfloor 0.0001 1.000000e-04
-topn 4 4
-topn_beam 0 0
-toprule
-transform legacy legacy
-unit_area yes yes
-upperf 6855.4976 6.855498e+03
-usewdphones no no
-uw 1.0 1.000000e+00
-var
-varfloor 0.0001 1.000000e-04
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wbeam 7e-29 7.000000e-29
-wip 0.65 6.500000e-01
-wlen 0.025625 2.562500e-02
INFO: cmd_ln.c(691): Parsing command line:
\
-nfilt 20 \
-lowerf 1 \
-upperf 4000 \
-wlen 0.025 \
-transform dct \
-round_filters no \
-remove_dc yes \
-svspec 0-12/13-25/26-38 \
-feat 1s_c_d_dd \
-agc none \
-cmn current \
-cmninit 56,-3,1 \
-varnorm no
Current configuration:
-agc none none
-agcthresh 2.0 2.000000e+00
-alpha 0.97 9.700000e-01
-ceplen 13 13
-cmn current current
-cmninit 8.0 56,-3,1
-dither no no
-doublebw no no
-feat 1s_c_d_dd 1s_c_d_dd
-frate 100 100
-input_endian little little
-lda
-ldadim 0 0
-lifter 0 0
-logspec no no
-lowerf 133.33334 1.000000e+00
-ncep 13 13
-nfft 512 512
-nfilt 40 20
-remove_dc no yes
-round_filters yes no
-samprate 16000 8.000000e+03
-seed -1 -1
-smoothspec no no
-svspec 0-12/13-25/26-38
-transform legacy dct
-unit_area yes yes
-upperf 6855.4976 4.000000e+03
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wlen 0.025625 2.500000e-02
INFO: acmod.c(242): Parsed model-specific feature parameters from
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/feat.params
INFO: feat.c(684): Initializing feature stream to type: '1s_c_d_dd',
ceplen=13, CMN='current', VARNORM='no', AGC='none'
INFO: cmn.c(142): mean= 12.00, mean= 0.0
INFO: acmod.c(163): Using subvector specification 0-12/13-25/26-38
INFO: mdef.c(520): Reading model definition:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/mdef
INFO: mdef.c(531): Found byte-order mark BMDF, assuming this is a binary mdef
file
INFO: bin_mdef.c(330): Reading binary model definition:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/mdef
INFO: bin_mdef.c(507): 50 CI-phone, 143047 CD-phone, 3 emitstate/phone, 150
CI-sen, 5150 Sen, 27135 Sen-Seq
INFO: tmat.c(205): Reading HMM transition probability matrices:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/transition_matrices
INFO: acmod.c(117): Attempting to use SCHMM computation module
INFO: ms_gauden.c(198): Reading mixture gaussian parameter:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/means
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(198): Reading mixture gaussian parameter:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/variances
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(354): 0 variance values floored
INFO: s2_semi_mgau.c(908): Loading senones from dump file
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/sendump
INFO: s2_semi_mgau.c(932): BEGIN FILE FORMAT DESCRIPTION
INFO: s2_semi_mgau.c(1027): Using memory-mapped I/O for senones
INFO: s2_semi_mgau.c(1304): Maximum top-N: 4 Top-N beams: 0 0 0
INFO: phone_loop_search.c(105): State beam -230231 Phone exit beam -115115
Insertion penalty 0
INFO: dict.c(306): Allocating 10346 * 20 bytes (202 KiB) for word entries
INFO: dict.c(321): Reading main dictionary:
/mnt/sdcard/edu.cmu.pocketsphinx/lm/en_US/hub4.5000.dic
INFO: dict.c(212): Allocated 44 KiB for strings, 69 KiB for phones
INFO: dict.c(324): 6212 words read
INFO: dict.c(330): Reading filler dictionary:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/noisedict
INFO: dict.c(212): Allocated 0 KiB for strings, 0 KiB for phones
INFO: dict.c(333): 38 words read
INFO: dict2pid.c(396): Building PID tables for dictionary
INFO: dict2pid.c(404): Allocating 50^3 * 2 bytes (244 KiB) for word-initial
triphones
INFO: dict2pid.c(131): Allocated 30200 bytes (29 KiB) for word-final triphones
INFO: dict2pid.c(195): Allocated 30200 bytes (29 KiB) for single-phone word
triphones
INFO: ngram_model_arpa.c(77): No \data\ mark in LM file
INFO: ngram_model_dmp.c(142): Will use memory-mapped I/O for LM file
INFO: ngram_model_dmp.c(196): ngrams 1=5001, 2=436879, 3=418286
INFO: ngram_model_dmp.c(242): 5001 = LM.unigrams(+trailer) read
INFO: ngram_model_dmp.c(291): 436879 = LM.bigrams(+trailer) read
INFO: ngram_model_dmp.c(317): 418286 = LM.trigrams read
INFO: ngram_model_dmp.c(342): 37293 = LM.prob2 entries read
INFO: ngram_model_dmp.c(362): 14370 = LM.bo_wt2 entries read
INFO: ngram_model_dmp.c(382): 36094 = LM.prob3 entries read
INFO: ngram_model_dmp.c(410): 854 = LM.tseg_base entries read
INFO: ngram_model_dmp.c(466): 5001 = ascii word strings read
INFO: ngram_search_fwdtree.c(99): 457 unique initial diphones
INFO: ngram_search_fwdtree.c(147): 0 root, 0 non-root channels, 54 single-
phone words
INFO: ngram_search_fwdtree.c(186): Creating search tree
INFO: ngram_search_fwdtree.c(191): before: 0 root, 0 non-root channels, 54
single-phone words
INFO: ngram_search_fwdtree.c(326): after: max nonroot chan increased to 13439
INFO: ngram_search_fwdtree.c(338): after: 457 root, 13311 non-root channels,
53 single-phone words
INFO: ngram_search_fwdflat.c(156): fwdflat: min_ef_width = 4, max_sf_win = 25
INFO: cmd_ln.c(691): Parsing command line:
Current configuration:
-agc none none
-agcthresh 2.0 2.000000e+00
-alpha 0.97 9.700000e-01
-ascale 20.0 2.000000e+01
-aw 1 1
-backtrace no no
-beam 1e-48 1.000000e-48
-bestpath yes yes
-bestpathlw 9.5 9.500000e+00
-bghist no no
-ceplen 13 13
-cmn current current
-cmninit 8.0 8.0
-compallsen no no
-debug 0
-dict
-dictcase no no
-dither no no
-doublebw no no
-ds 1 1
-fdict
-feat 1s_c_d_dd 1s_c_d_dd
-featparams
-fillprob 1e-8 1.000000e-08
-frate 100 100
-fsg
-fsgusealtpron yes yes
-fsgusefiller yes yes
-fwdflat yes yes
-fwdflatbeam 1e-64 1.000000e-64
-fwdflatefwid 4 4
-fwdflatlw 8.5 8.500000e+00
-fwdflatsfwin 25 25
-fwdflatwbeam 7e-29 7.000000e-29
-fwdtree yes yes
-hmm
-input_endian little little
-jsgf
-kdmaxbbi -1 -1
-kdmaxdepth 0 0
-kdtree
-latsize 5000 5000
-lda
-ldadim 0 0
-lextreedump 0 0
-lifter 0 0
-lm
-lmctl
-lmname default default
-logbase 1.0001 1.000100e+00
-logfn
-logspec no no
-lowerf 133.33334 1.333333e+02
-lpbeam 1e-40 1.000000e-40
-lponlybeam 7e-29 7.000000e-29
-lw 6.5 6.500000e+00
-maxhmmpf -1 -1
-maxnewoov 20 20
-maxwpf -1 -1
-mdef
-mean
-mfclogdir
-min_endfr 0 0
-mixw
-mixwfloor 0.0000001 1.000000e-07
-mllr
-mmap yes yes
-ncep 13 13
-nfft 512 512
-nfilt 40 40
-nwpen 1.0 1.000000e+00
-pbeam 1e-48 1.000000e-48
-pip 1.0 1.000000e+00
-pl_beam 1e-10 1.000000e-10
-pl_pbeam 1e-5 1.000000e-05
-pl_window 0 0
-rawlogdir
-remove_dc no no
-round_filters yes yes
-samprate 16000 1.600000e+04
-seed -1 -1
-sendump
-senlogdir
-senmgau
-silprob 0.005 5.000000e-03
-smoothspec no no
-svspec
-tmat
-tmatfloor 0.0001 1.000000e-04
-topn 4 4
-topn_beam 0 0
-toprule
-transform legacy legacy
-unit_area yes yes
-upperf 6855.4976 6.855498e+03
-usewdphones no no
-uw 1.0 1.000000e+00
-var
-varfloor 0.0001 1.000000e-04
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wbeam 7e-29 7.000000e-29
-wip 0.65 6.500000e-01
-wlen 0.025625 2.562500e-02
INFO: cmd_ln.c(691): Parsing command line:
\
-nfilt 20 \
-lowerf 1 \
-upperf 4000 \
-wlen 0.025 \
-transform dct \
-round_filters no \
-remove_dc yes \
-svspec 0-12/13-25/26-38 \
-feat 1s_c_d_dd \
-agc none \
-cmn current \
-cmninit 56,-3,1 \
-varnorm no
Current configuration:
-agc none none
-agcthresh 2.0 2.000000e+00
-alpha 0.97 9.700000e-01
-ceplen 13 13
-cmn current current
-cmninit 8.0 56,-3,1
-dither no no
-doublebw no no
-feat 1s_c_d_dd 1s_c_d_dd
-frate 100 100
-input_endian little little
-lda
-ldadim 0 0
-lifter 0 0
-logspec no no
-lowerf 133.33334 1.000000e+00
-ncep 13 13
-nfft 512 512
-nfilt 40 20
-remove_dc no yes
-round_filters yes no
-samprate 16000 8.000000e+03
-seed -1 -1
-smoothspec no no
-svspec 0-12/13-25/26-38
-transform legacy dct
-unit_area yes yes
-upperf 6855.4976 4.000000e+03
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wlen 0.025625 2.500000e-02
INFO: acmod.c(242): Parsed model-specific feature parameters from
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/feat.params
INFO: feat.c(684): Initializing feature stream to type: '1s_c_d_dd',
ceplen=13, CMN='current', VARNORM='no', AGC='none'
INFO: cmn.c(142): mean= 12.00, mean= 0.0
INFO: acmod.c(163): Using subvector specification 0-12/13-25/26-38
INFO: mdef.c(520): Reading model definition:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/mdef
INFO: mdef.c(531): Found byte-order mark BMDF, assuming this is a binary mdef
file
INFO: bin_mdef.c(330): Reading binary model definition:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/mdef
INFO: bin_mdef.c(507): 50 CI-phone, 143047 CD-phone, 3 emitstate/phone, 150
CI-sen, 5150 Sen, 27135 Sen-Seq
INFO: tmat.c(205): Reading HMM transition probability matrices:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/transition_matrices
INFO: acmod.c(117): Attempting to use SCHMM computation module
INFO: ms_gauden.c(198): Reading mixture gaussian parameter:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/means
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(198): Reading mixture gaussian parameter:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/variances
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(294): 256x13
INFO: ms_gauden.c(354): 0 variance values floored
INFO: s2_semi_mgau.c(908): Loading senones from dump file
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/sendump
INFO: s2_semi_mgau.c(932): BEGIN FILE FORMAT DESCRIPTION
INFO: s2_semi_mgau.c(1027): Using memory-mapped I/O for senones
INFO: s2_semi_mgau.c(1304): Maximum top-N: 4 Top-N beams: 0 0 0
INFO: phone_loop_search.c(105): State beam -230231 Phone exit beam -115115
Insertion penalty 0
INFO: dict.c(306): Allocating 10346 * 20 bytes (202 KiB) for word entries
INFO: dict.c(321): Reading main dictionary:
/mnt/sdcard/edu.cmu.pocketsphinx/lm/en_US/hub4.5000.dic
INFO: dict.c(212): Allocated 44 KiB for strings, 69 KiB for phones
INFO: dict.c(324): 6212 words read
INFO: dict.c(330): Reading filler dictionary:
/mnt/sdcard/edu.cmu.pocketsphinx/hmm/en_US/hub4wsj_sc_8k/noisedict
INFO: dict.c(212): Allocated 0 KiB for strings, 0 KiB for phones
INFO: dict.c(333): 38 words read
INFO: dict2pid.c(396): Building PID tables for dictionary
INFO: dict2pid.c(404): Allocating 50^3 * 2 bytes (244 KiB) for word-initial
triphones
INFO: dict2pid.c(131): Allocated 30200 bytes (29 KiB) for word-final triphones
INFO: dict2pid.c(195): Allocated 30200 bytes (29 KiB) for single-phone word
triphones
INFO: ngram_model_arpa.c(77): No \data\ mark in LM file
INFO: ngram_model_dmp.c(142): Will use memory-mapped I/O for LM file
INFO: ngram_model_dmp.c(196): ngrams 1=5001, 2=436879, 3=418286
INFO: ngram_model_dmp.c(242): 5001 = LM.unigrams(+trailer) read
INFO: ngram_model_dmp.c(291): 436879 = LM.bigrams(+trailer) read
INFO: ngram_model_dmp.c(317): 418286 = LM.trigrams read
INFO: ngram_model_dmp.c(342): 37293 = LM.prob2 entries read
INFO: ngram_model_dmp.c(362): 14370 = LM.bo_wt2 entries read
INFO: ngram_model_dmp.c(382): 36094 = LM.prob3 entries read
INFO: ngram_model_dmp.c(410): 854 = LM.tseg_base entries read
INFO: ngram_model_dmp.c(466): 5001 = ascii word strings read
INFO: ngram_search_fwdtree.c(99): 457 unique initial diphones
INFO: ngram_search_fwdtree.c(147): 0 root, 0 non-root channels, 54 single-
phone words
INFO: ngram_search_fwdtree.c(186): Creating search tree
INFO: ngram_search_fwdtree.c(191): before: 0 root, 0 non-root channels, 54
single-phone words
INFO: ngram_search_fwdtree.c(326): after: max nonroot chan increased to 13439
INFO: ngram_search_fwdtree.c(338): after: 457 root, 13311 non-root channels,
53 single-phone words
INFO: ngram_search_fwdflat.c(156): fwdflat: min_ef_width = 4, max_sf_win = 25
that's the example for english recognition and for french it doesn't display
anything
i'll try to do the process of french recognition on desktop .
thanks .
This issue is covered in FAQ:
http://cmusphinx.sourceforge.net/wiki/faq#qwhy_my_accuracy_is_poor
Ok, share the results.
ok .
I've seen this question in FAQ but that mean in my case that there is a
specifique configuration to do for the decoder because my accuracy is very low
. In this case how can i do this please ? is there any tuto for this ?
Nothing like that is written there, check the entry again
Just do what FAQ answer suggest
the first sentence in the answer is :
" The first thing you need to understand if your accuracy just lower than
expected or very low. If it's very low most likely you misconfigured the
decoder. "