Skip to content

Commit

Permalink
Browse files Browse the repository at this point in the history
finished updating tool documents
  • Loading branch information
LeeAkinobu committed Jan 24, 2019
1 parent 2c817e1 commit c0a3a21
Show file tree
Hide file tree
Showing 15 changed files with 282 additions and 98 deletions.
11 changes: 9 additions & 2 deletions adinrec/README.md
Expand Up @@ -17,7 +17,7 @@ Record one utterance from audio device and save to a file.
`adinrec` detects an utterance input and store it to a file.

This tool uses Julius's internal VAD module for speech detection. The detection
algorithm and parameters are the same as Julius.
algorithm and parameters are the same as Julius.

The audio format is 16 bit, 1 channel in Microsoft WAV format. If the given
filename already exists, it will be overridden. When filename is "-" , the
Expand Down Expand Up @@ -90,7 +90,14 @@ Input latency of microphone input in milliseconds. Smaller value will shorten
latency but sometimes make process unstable. Default value will depend on the
running OS.

## Related tools

- "[adintool](https://github.com/julius-speech/julius/tree/master/adintool)" is
another recording software with rich functions.
- "[julius](https://github.com/julius-speech/julius/)" uses the same audio
detection algorithm.

## License

This tool is licensed under the same license with Julius. See the license term
of Julius for details.
of Julius for details.
9 changes: 8 additions & 1 deletion adintool/README.md
Expand Up @@ -47,7 +47,7 @@ Output waveform / feature vector:
- none

This tool uses Julius's internal VAD module for speech detection. The detection
algorithm and parameters are the same as Julius.
algorithm and parameters are the same as Julius.

The default audio format is 16 bit, 1 channel in Microsoft WAV format.

Expand Down Expand Up @@ -233,6 +233,13 @@ Input latency of microphone input in milliseconds. Smaller value will shorten
latency but sometimes make process unstable. Default value will depend on the
running OS.

## Related tools

- "[adinrec](https://github.com/julius-speech/julius/tree/master/adinrec)" is a
simplified version recording only the first audio segment.
- "[julius](https://github.com/julius-speech/julius/)" uses the same audio
detection algorithm, and can receive data from `adintool`.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
9 changes: 6 additions & 3 deletions binlm2arpa/README.md
Expand Up @@ -42,11 +42,14 @@ Convert forward 5-gram binlm into ARPA file:
writing forward 5-gram to "outputFile.ngram.arpa"
```

## Options
## Related tools

No option.
- "[mkbingram](https://github.com/julius-speech/julius/tree/master/mkbingram)"
can convert ARPA files into binary N-gram.
- "[generate-ngram](https://github.com/julius-speech/julius/tree/master/generate-ngram)"
can generate random sentences from binary N-gram.

## License

This tool is licensed under the same license with Julius. See the license term
of Julius for details.
of Julius for details.
7 changes: 7 additions & 0 deletions generate-ngram/README.md
Expand Up @@ -85,6 +85,13 @@ Verbose output.

Debug output.

## Related tools

- "[mkbingram](https://github.com/julius-speech/julius/tree/master/mkbingram)"
can convert ARPA files into binary N-gram.
- "[binlm2arpa](https://github.com/julius-speech/julius/tree/master/binlm2arpa)"
is can revert binary N-gram into its original ARPA format.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
11 changes: 11 additions & 0 deletions gramtools/accept_check/README.md
Expand Up @@ -66,6 +66,17 @@ Short-pause word name to be skipped. (default: "sp")

Debug output.

## Related tools

- "[nextword](https://github.com/julius-speech/julius/tree/master/gramtools/nextword)"
can show word prediction of a grammar at given context for debug.
- "[generate](https://github.com/julius-speech/julius/tree/master/gramtools/generate)"
can generate sentences randomly according to the grammar.
- "[mkdfa.pl](https://github.com/julius-speech/julius/tree/master/gramtools/mkdfa)"
is the grammar compiler for Julius.
- "[yomi2voca.pl](https://github.com/julius-speech/julius/tree/master/gramtools/yomi2voca)"
can convert Japanese Hiragana to phoneme for making recognition dictionary.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
5 changes: 5 additions & 0 deletions gramtools/dfa_determinize/README.md
Expand Up @@ -36,6 +36,11 @@ Determinizing a non-deterministic grammar automaton into deterministic form:

Specify output file. (default: output to standard output)

## Related tools

- "[dfa_minimize](https://github.com/julius-speech/julius/tree/master/gramtools/dfa_minimize)"
can minimize DFA.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
8 changes: 8 additions & 0 deletions gramtools/dfa_minimize/README.md
Expand Up @@ -37,6 +37,14 @@ Minimize a dfa file:

Specify output file. (default: output to standard output)

## Related tools

- "[mkdfa.pl](https://github.com/julius-speech/julius/tree/master/gramtools/mkdfa)"
is the grammar compiler for Julius. `dfa_minimize` is called within this
script.
- "[dfa_determinize](https://github.com/julius-speech/julius/tree/master/gramtools/dfa_determinize)"
can determinize DFA.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
9 changes: 9 additions & 0 deletions gramtools/generate/README.md
Expand Up @@ -72,6 +72,15 @@ Short-pause word name to be suppressed. (default: "sp")

Debug output.

## Related tools

- "[accept_check](https://github.com/julius-speech/julius/tree/master/gramtools/accept_check)"
can check sentence acceptance.
- "[nextword](https://github.com/julius-speech/julius/tree/master/gramtools/nextword)"
can show word prediction of a grammar at given context for debug.
- "[mkdfa.pl](https://github.com/julius-speech/julius/tree/master/gramtools/mkdfa)"
is the grammar compiler for Julius.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
42 changes: 34 additions & 8 deletions gramtools/mkdfa/README.md
Expand Up @@ -10,22 +10,33 @@ Grammar compiler for Julius.

## Description

`mkdfa.pl` compiles Julius grammar definition (BNF-style definition file `.grammar` and vocabulary file `.voca`) into finite automaton grammar to be used in Julius.
`mkdfa.pl` compiles Julius grammar definition (BNF-style definition file
`.grammar` and vocabulary file `.voca`) into finite automaton grammar to be
used in Julius.

The output files, `prefix.dfa` and `prefix.dict`, can be read by Julius to perform grammar-based recognition based on the given grammar.
The output files, `prefix.dfa` and `prefix.dict`, can be read by Julius to
perform grammar-based recognition based on the given grammar.

### Prerequisites

See [Julius grammar-kit GitHub](https://github.com/julius-speech/grammar-kit/) for details about grammars. There are also an [example](https://github.com/julius-speech/grammar-kit/tree/master/SampleGrammars_en).
Perl is required to run this program. Also put `mkfa` and `dfa_minimize` in the
same directory of `mkdfa.pl`.

See [Julius grammar-kit GitHub](https://github.com/julius-speech/grammar-kit/)
for details about grammars. There are also an
[example](https://github.com/julius-speech/grammar-kit/tree/master/SampleGrammars_en).

### Installing

This tool will be installed together with Julius.
Place `mkfa` and `dfa_minimize` at the same directory as `mkdfa.pl`, since it invokes them internally while compilation.
This tool will be installed together with Julius. Place `mkfa` and
`dfa_minimize` at the same directory as `mkdfa.pl`, since it invokes them
internally while compilation.

## Usage

Compiling the [sample grammar](https://github.com/julius-speech/grammar-kit/tree/master/SampleGrammars_en) "fruit":
Compiling the [sample
grammar](https://github.com/julius-speech/grammar-kit/tree/master/SampleGrammars_en)
"fruit":

```shell
% git clone https://github.com/julius-speech/grammar-kit/
Expand Down Expand Up @@ -62,19 +73,34 @@ or

### `-n`

Skip voca-to-dict conversion. Build `.dfa` from `.grammar` but leave `.dict` unchanged.
Skip voca-to-dict conversion. Build `.dfa` from `.grammar` but leave `.dict`
unchanged.

## Environment Variables

### TMP / TEMP

Temporary directory. When not specified, one from the following list will be used:
Temporary directory. When not specified, one from the following list will be
used:

- /tmp
- /var/tmp
- /WINDOWS/Temp
- /WINNT/Temp.

## Related tools

- "[dfa_minimize](https://github.com/julius-speech/julius/tree/master/gramtools/dfa_minimize)"
can minimize DFA. `mkdfa.pl` call call this inside the process.
- - "[mkfa](https://github.com/julius-speech/julius/tree/master/gramtools/mkfa)"
is the core compiler. `mkdfa.pl` call call this inside the process.
- "[generate](https://github.com/julius-speech/julius/tree/master/gramtools/generate)"
can generate sentences randomly according to the grammar.
- "[accept_check](https://github.com/julius-speech/julius/tree/master/gramtools/accept_check)"
can check sentence acceptance.
- "[nextword](https://github.com/julius-speech/julius/tree/master/gramtools/nextword)"
can show word prediction of a grammar at given context for debug.

## License

This tool is licensed under the same license with Julius. See the license term
Expand Down
81 changes: 0 additions & 81 deletions gramtools/nextword/00readme.txt

This file was deleted.

0 comments on commit c0a3a21

Please sign in to comment.