# Julius (software)

> Mediated Wiki article. Canonical URL: https://mediated.wiki/source/Julius_(software)
> Markdown URL: https://mediated.wiki/source/Julius_(software).md
> Source: https://en.wikipedia.org/wiki/Julius_(software)
> Source revision: 1263306675
> License: Creative Commons Attribution-ShareAlike 4.0 International (https://creativecommons.org/licenses/by-sa/4.0/)

{{primary sources|date=April 2015}}
{{Infobox software
| name = Julius
| logo = <!-- Image name only -->
| logo alt = 
| logo caption = 
| screenshot = <!-- Image name only -->
| screenshot alt = 
| caption = 
| author = Lee Akinobu
| developer = Kawahara Lab., Kyoto University<br />Julius project team, [Nagoya Institute of Technology](/source/Nagoya_Institute_of_Technology)
| released = {{Start date and age|1991}}
| latest release version = {{#statements:software version identifier}}
| latest release date = {{wikidata|qualifier|P348|P577}}
| repo = {{URL|github.com/julius-speech}}
| programming language = [C](/source/C_(programming_language))
| operating system = [Unix](/source/Unix) ([Linux](/source/Linux), [BSD](/source/Berkeley_Software_Distribution), etc.), [Windows](/source/Microsoft_Windows) (via [Cygwin](/source/Cygwin))
| platform = [IA-32](/source/IA-32), [x86-64](/source/x86-64)
| size = 
| language = Japanese, English
| language count = <!-- Number only -->
| genre = [Speech recognition](/source/Speech_recognition)
| license = [Free](/source/Free_software), [BSD](/source/BSD_licenses) style<ref>{{cite web |url=https://fedoraproject.org/wiki/Licensing/Julius |title=Licensing/Julius |last=Callaway |first=Tom (spot) |date=2012-08-13 |website=Fedora Wiki |publisher=Red Hat |access-date=2019-03-24}}</ref><ref name="Lizenz">{{cite web |url=http://julius.osdn.jp/license.txt |title=Large Vocabulary Continuous Speech Recognition Engine Julius |author=<!--Unstated--> |date=2014 |website=Julius development team |publisher=Nagoya Institute of Technology |access-date=2019-03-24 |archive-date=2019-08-03 |archive-url=https://web.archive.org/web/20190803022127/http://julius.osdn.jp/LICENSE.txt |url-status=dead }}</ref>
| website = {{URL|julius.osdn.jp/en_index.php}}
}}

'''Julius''' is a [speech recognition](/source/speech_recognition) engine, specifically a high-performance, two-pass large vocabulary continuous speech recognition (LVCSR) decoder software for speech-related researchers and developers. It can perform almost [real-time computing](/source/real-time_computing) (RTC) decoding on most current [personal computer](/source/personal_computer)s (PCs) in 60k word dictation task using word [trigram](/source/trigram) (3-gram) and context-dependent [Hidden Markov model](/source/Hidden_Markov_model) (HMM). Major search methods are fully incorporated.

It is also modularized carefully to be independent from model structures, and various HMM types are supported such as shared-state [triphone](/source/triphone)s and tied-mixture models, with any number of mixtures, states, or phones. Standard formats are adopted to cope with other free modeling toolkit. The main platform is [Linux](/source/Linux) and other [Unix](/source/Unix) workstations, and it works on [Windows](/source/Microsoft_Windows). Julius is [free and open-source software](/source/free_and_open-source_software), released under a revised [BSD](/source/BSD_licenses) style [software license](/source/software_license).

Julius has been developed as part of a free software toolkit for Japanese LVCSR research since 1997, and the work has been continued at Continuous Speech Recognition Consortium (CSRC), Japan from 2000 to 2003.

From rev.3.4, a grammar-based recognition parser named ''Julian'' is integrated into Julius. Julian is a modified version of Julius that uses hand-designed type of [finite-state machine](/source/finite-state_machine) (FSM) termed a [deterministic finite automaton](/source/deterministic_finite_automaton) (DFA) grammar as a language model. It can be used to build a kind of voice command system of small vocabulary, or various spoken [dialog system](/source/dialog_system) tasks.

==About models==
To run, the Julius recognizer needs a [language model](/source/language_model) and an [acoustic model](/source/acoustic_model) for each language.

Julius adopts acoustic models in Hidden Markov Model Toolkit ([HTK](/source/HTK_(software))) [ASCII](/source/ASCII) format, pronunciation dictionary in HTK-like format, and word 3-gram language models in ARPA standard format: forward 2-gram and reverse 3-gram as trained from [speech corpus](/source/speech_corpus) with reversed word order.

Although Julius is only distributed with Japanese models, the [VoxForge](/source/VoxForge) project is working to create English acoustic models for use with the Julius Speech Recognition Engine.

In April 2018, thanks to the effort of [Mozilla](/source/Mozilla) foundation, a 350-hour audio corpus of spoken English was made available. The new English ENVR-v5.4 open-source speech model was released along with Polish PLPL-v7.1 models and are available from SourceForge.<ref>{{cite web | url=https://sourceforge.net/projects/juliusmodels/files/ | title=JuliusModels - Browse Files at SourceForge.net }}</ref>

==See also==
* [List of speech recognition software](/source/List_of_speech_recognition_software)

==References==
{{Reflist}}

==External links==
* {{Official website|julius.osdn.jp/en_index.php}}, at osdn.jp

Category:Speech recognition software
Category:Computational linguistics
Category:Free software projects

{{Comp-ling-stub}}

---
Adapted from the Wikipedia article [Julius (software)](https://en.wikipedia.org/wiki/Julius_(software)) by Wikipedia contributors ([contributor history](https://en.wikipedia.org/wiki/Julius_(software)?action=history)). Available under [Creative Commons Attribution-ShareAlike 4.0 International](https://creativecommons.org/licenses/by-sa/4.0/). Changes may have been made.