Hello,
My AM limited with 60 words is organized in uppercase..
DIC: ABORT AH B AO R T ACCELERATE AE K S EH L ER EY T ACTIVATE AE K T AH V EY T
TRANSCRIPT:
<s> ++NOISE++ ++BREATH++ LEFT DOORS OPEN RIGHT DOORS OPEN RIGHT DOORS OPEN ALL DOORS OPEN ALL DOORS OPEN LEFT CLOSE LEFT CLOSE RIGHT CLOSE RIGHT CLOSE ALL CLOSE ++BREATH++ ++NOISE++ </s> (peter_42_46)
If i need to adopt with the EN_US acoustic model, Do i need to change all the transcript and dictionary to Lowercase??
Do i need to change all the transcript and dictionary to Lowercase??
No
I am having trouble in testing the Adapter model, errors occured in dict.c andf batch.c files
I used the commands mentioned in adapt.sh , Here are all the files to reproduce the result
https://www.dropbox.com/s/hun81ls40t51yam/adapt2.tar.gz?dl=0
The error snapshot is as follows
... ... ...
ERROR: "dict.c", line 205: Line 133433: Failed to add the word 'zyuganov(2)' (duplicate?); ignored ERROR: "dict.c", line 107: Missing base word for: zyuganov's(2) ERROR: "dict.c", line 205: Line 133434: Failed to add the word 'zyuganov's(2)' (duplicate?); ignored INFO: dict.c(213): Allocated 907 KiB for strings, 1537 KiB for phones INFO: dict.c(336): 124207 words read INFO: dict.c(342): Reading filler dictionary: en-us-adapt/noisedict ERROR: "dict.c", line 205: Line 1: Failed to add the word '<s>' (duplicate?); ignored ERROR: "dict.c", line 205: Line 2: Failed to add the word '</s>' (duplicate?); ignored ERROR: "dict.c", line 205: Line 3: Failed to add the word '<sil>' (duplicate?); ignored ERROR: "dict.c", line 205: Line 4: Failed to add the word '<SIL>' (duplicate?); ignored ERROR: "dict.c", line 195: Line 5: Phone '+NOISE+' is mising in the acoustic model; word '++noise++' ignored ERROR: "dict.c", line 195: Line 6: Phone '+BREATH+' is mising in the acoustic model; word '++breath++' ignored INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones INFO: dict.c(345): 0 words read ERROR: "dict.c", line 371: Word '<sil>' must occur (only) in filler dictionary FATAL: "batch.c", line 820: PocketSphinx decoder init failed
Log in to post a comment.
Hello,
My AM limited with 60 words is organized in uppercase..
DIC:
ABORT AH B AO R T
ACCELERATE AE K S EH L ER EY T
ACTIVATE AE K T AH V EY T
TRANSCRIPT:
If i need to adopt with the EN_US acoustic model, Do i need to change all the transcript and dictionary to Lowercase??
Last edit: Nayak BS 2015-04-19
No
I am having trouble in testing the Adapter model, errors occured in dict.c andf batch.c files
I used the commands mentioned in adapt.sh , Here are all the files to reproduce the result
https://www.dropbox.com/s/hun81ls40t51yam/adapt2.tar.gz?dl=0
Last edit: Nayak BS 2015-04-20
The error snapshot is as follows
...
...
...