name: isip_make_db

synopsis:

isip_make_db [options] [audio_db.sof] [trans_db.sof]

quick start:


isip_make_db -help;

isip_make_db -db audio -audio audio_list.text -name TIDIGITS -type text audio_db_text.sof

isip_make_db -db transcription -transcription trans.text -level word -name TIDIGITS \
-type text trans_db_text.sof

description:

options:

-audio : list of audio files with identifiers in text
-db: database type (default=audio, transcription, both)
-debug: debug level (default=NONE, BRIEF, DETAILED, ALL)
-help: display a help message
-level: transcription level
-log_file: log messages
-name: name of the database
-transcription: list of transcriptions with identifiers in text
-type: output database type(default=text, binary)

arguments:

audio_db.sof: output audio database
trans_db.sof: output transcription database

formats:

audio.text: filename [identifier]
trans.text: trans or
                   identifier: trans or
                   identifier start_time stop_time: trans or
                   identifier start_time stop_time channel: trans or
                   : trans1 : trans2 :trans3 : trans4

examples:

This example shows how to create an audio database using the input audio list that contains the audio filenames and the corresponding identifiers in the format described above.
```
isip_make_db -db audio -audio audio_list_0.text -name TIDIGITS -type text \

audio_db_text_0.sof
```
If the identifiers are not provided in the input audio list, the line number of the audio filenames in the input audio list is extracted as the identifiers.
```
isip_make_db -db audio -audio audio_list_1.text -name TIDIGITS -type text \

audio_db_text_1.sof
```
This example shows how to create a transcription database given the identifiers, start-time of the utterances in the audio database, stop-time of the utterances in the audio database, the channel numbers and, the transcription itself, as the input transcription file in the format specified above.
```
isip_make_db -db transcription -transcription trans_0.text -name TIDIGITS -level word \
-type text trans_db_text_0.sof
```
The identifiers, start-time, stop-time and the channel number are optional. If the identifiers are not specified in the input transcription file, the line number of the transcription in the input file is extracted as the identifiers. If the start-time and the stop-time are not specified in the input transcription file, the anchors on the output transcription database are flagged as false.
```
isip_make_db -db transcription -transcription trans_1.text -name TIDIGITS -level word \
-type text trans_db_text_1.sof
```
The same example shown above without identifiers, start-time, stop-time, and, the channel number can also be specified in a different format. In this format, all the transcriptions are specified in the same line, each beginning with a ":".
```
isip_make_db -db transcription -transcription trans_2.text -name TIDIGITS -level word \
-type text trans_db_text_2.sof
```

This example shows how to create the audio database and the corresponding transcription database.


isip_make_db -db both -audio audio_list_0.text -transcription trans_0.text -name TIDIGITS \
-level word -type text audio_db_text_0.sof trans_db_text_0.sof

manuals:

None. languages:

This utility is written entirely in C++. environment variables:

None. files:

./isip_make_db.h: header file
./isip_make_db.cc: main file
./usage_message.text: usage message
./help_message.text: help message

see also:

notes:

none