Atype: Persian Voice Dictation

Atype is the first Persian (Farsi) speech dictation software designed and developed based on a large vocabulary, speaker independent speech recognition engine. Using this package, you do not need to type by keyboard, just talk to microphone and Atype converts it to text.

Nearly all computer users do typing; managers for writing letters, notes and reports, clerks for preparing organizational and administrative reports, doctors for typing medical reports, lawyers for preparing legal documents, personal users for writing notes, diaries, students for writing papers, theses, notes, etc. An essential problem of most computer users is that they are not proficient in typing. As a result, manual typing turns into a time-consuming hassle. Thanks to the improvements in speech recognition technology, this concern has been resolved to a large extent.

Speech recognition facilitates typing, saves considerable time and consequently reduces costs. Persian (Farsi) speech is recognized instantly without dependence on speaker. Thus, you can save time and cost significantly in writing documents in Persian. The flexibility of Atype in adverse environmental conditions is achieved by advanced noise robustness techniques. The background adaptation makes the recognition more accurate as the time goes on.

Atype is a continuous speech recognition engine for Persian that is customized for various applications. There are various versions of Atype that can be installed as a single-user package or a multi-user network service.

 

Atype Versions

Atype with a continuous speech recognition engine is speaker independent and contains a large vocabulary. Using this engine, ALBORZ TECH has different versions to offer each with its own capabilities and for specific applications. Atype is available in single and enterprise (on network) versions. Atype versions include:

 

  • Atype-Professional: specialized for organizations and agencies, typists with a high diversity of typing tasks
  • Atype-Medical: to prepare radiological reports such as MRI, CT, etc.
  • Atype-Legal: judicial and legal texts

 

Atype versions represent the state-of-the-art in Persian large-vocabulary, speaker-independent speech recognition. Our current researches continue to improve the state of our speech recognition technology and deliver significant improvements in recognition accuracy for spontaneous speech in meeting environments.

 

Atype Features

  • High accuracy (up to 95%)
  • Real-time typing, it types as you talk
  • Working in all typing environments such as M.S. Word, forms, automation systems, etc.
  • Writing numbers and punctuations such as points and question marks by talking
  • Converting pre-recorded audio files into text (offline mode)
  • Containing a large number of words and phrases of Persian language
  • Capability to add new words to the vocabulary by the user (user lexicon)
  • Ability to customize vocabulary for specific applications
  • Speaker adaptation capability to learn the speakers’ accent and tone (user profile)
  • Automatic speaker recognition ability to select user profile automatically (Avoice service)
  • Persian text-to-speech ability to read the typed text (Atalk service)
  • Environmental adaptation capability to maintain performance in noisy conditions
  • Supports voice commands to run and control applications (Atask service)
  • Providing record history for the next editions of texts
  • A customized editor with special facilities such as Atext service
  • Coming with a special speech recognition headset
  • Bilingual user interface (English and Persian)
  • Having single-user and multi-user network versions

 

 

Atype users

Because of the widespread usage of Atype and its user-friendly features, this software encompasses a wide range of users. Some Atype users are:

  • Managers, experts, heads of offices, secretaries and typists in all organizations, banking and financial institutions, ministries, companies and governmental and non-governmental agencies, typing offices, …
  • Physicians in hospitals, radiology centers, MRI, CT and private offices for writing various medical reports
  • lawyers in legal offices, notaries, courts and …
  • Personal and home users, journalists, writers, university professors and …

 

Atype Benefits

  • Creating documents easily

Saving time and increasing the speed of typing and data insertion up to 3 times. The average speed of talking (or typing by speech), is about 105 words per minute, while the average speed of typing in a normal person, is about 33 words per minute.

  • Reduction in personal and organizational costs
  • Increasing the speed of responding to customers at offices, organizations, medical centers and legal centers
    security and confidentiality of data during data insertion (for managers who want to type their own secret texts)
  • Reducing work volume and creating the energy to encourage all staff to produce their own documents.
  • Preventing misspellings in typing documentation
  • Help the blind and the disabled to work and communicate with the computer

 

 

  • Atype-Professional

Atype-Professional is a large-vocabulary Persian speech dictation system designed for voice typing of general texts having high diversity of words such as in organizations and agencies. The vocabulary size of 65,000 Persian words has allowed Atype-Professional to cover most frequent words in organization letters and reports. In addition to high accuracy, other exclusive features of this package are typing the contents of audio files, supporting voice commands and automatic selection of user profiles. The complete feature lists of Atype-Professional are:

 

Atype-Professional Features

  • High accuracy (up to 95%)
  • Real-time typing, it types as you talk
  • Working in all typing environments such as M.S. Word, forms, automation systems, etc.
  • Converting pre-recorded audio files into text (offline mode)
  • Containing 65,000 most frequent Persian words in general applications
  • Capability to add new words to the vocabulary by the user (user lexicon)
  • Speaker adaptation capability to learn the speakers’ accent and tone (user profile, unlimited)
  • Automatic speaker recognition ability to select user profile automatically (Avoice service)
  • Persian text-to-speech ability to read the typed text (Atalk service)
  • Environmental adaptation capability to maintain performance in noisy conditions
  • Supports voice commands to run and control applications (Atask service)
  • Providing record history for the next editions of texts
  • Writing numbers and punctuations such as points and question marks by talking
  • A customized editor with special facilities such as Atext service
  • Coming with a special speech recognition headset
  • Bilingual user interface (English and Persian)
  • Having single-user and multi-user network versions

 

Atype-Professional Package includes:

  • Software CD
  • Atype user guide CD
  • Speech recognition microphone
  • USB lock
  • User manual booklet

 

Atype-Professional requirements

  • Windows XP operating system or higher
  • Installation of Persian Language and keyboard
  • Dot Net 3.0 Environment (to be installed along with the software)
  • 2 GHz CPU or higher
  • 1GB of RAM for Windows XP and 1.5 GB for Windows Vista and Win7
  • 300 MB of free disk space
  • Sound card with microphone input
  • High quality speech recognition microphone (to be supplied with software)

 

 

  • Atype-Medical

Atype-Medical is a customized version of Atype Persian voice dictation system for medical applications in radiology, MRI and CT. Using this package, medicine and radiologists are able to prepare their reports in a short time and reduce their costs. Atype-Medical can be deployed in all hospitals, treatment centers, radiology centers, MRI and CTs. The lexicon, language models and acoustic models of Atype are customized to results in high performance if the medical context. The list of Atype-Medical features is given in the following.

 

Atype-Medical Features

  • High accuracy (up to 95%)
  • Real-time typing, it types as you talk
  • Working in all typing environments such as M.S. Word, forms, automation systems, etc.
  • Converting pre-recorded audio files into text (offline mode)
  • Containing more than 30.000 words of radiology, MRI and CT
  • Containing Latin terms and expressions in medical field and radiology
  • Capability to add new words to the vocabulary by the user (user lexicon)
  • Speaker adaptation capability to learn the speakers’ accent and tone (user profile, unlimited)
  • Automatic speaker recognition ability to select user profile automatically (Avoice service)
  • Persian text-to-speech ability to read the typed text (Atalk service)
  • Environmental adaptation capability to maintain performance in noisy conditions
  • Supports voice commands to run and control applications (Atask service)
  • Providing record history for upcoming edition by physician or secretary
  • Writing numbers and punctuations such as points and question marks by talking
  • A customized editor with special facilities such as Atext service
  • Coming with a special speech recognition headset
  • Bilingual user interface (English and Persian)
  • Having single-user and multi-user network versions

 

Atype- Medical Package includes:

  • Software CD
  • Atype user guide CD
  • Speech recognition microphone
  • USB lock
  • User manual booklet

 

Atype- Medical requirements

  • Windows XP operating system or higher
  • Installation of Persian Language and keyboard
  • Dot Net 3.0 Environment (to be installed along with the software)
  • 2 GHz CPU or higher
  • 512 MB of RAM for Windows XP and 1 GB for Windows Vista and Win7
  • 250 MB of free disk space
  • Sound card with microphone input
  • High quality speech recognition microphone (to be supplied with software)

 

  • Atype-Legal

Atype-Legal is a customized version of Atype Persian voice dictation system that is developed for legal and judicial texts. Atype–Legal is suitable for attorneys, lawyers, judges, notaries, courts typists, etc. By using Atype-Legal, writing legal texts and documents is done quickly and easily with lower cost. In addition to useful common words in Persian, Atype-Legal vocabulary contains specific legal and judicial words. Moreover, Atype-Legal engine suits the style of legal writing. Capabilities of Atype-legal are:

 

Atype-Legal Features

  • High accuracy (up to 95%)
  • Real-time typing, it types as you talk
  • Working in all typing environments such as M.S. Word, forms, automation systems, etc.
  • Converting pre-recorded audio files into text (offline mode)
  • Containing vocabulary bank of more than 40,000 Persian words and phrases (general and specialized in legal and judicial field)
  • Containing Latin terms and expressions in medical field and radiology
  • Capability to add new words to the vocabulary by the user (user lexicon)
  • Speaker adaptation capability to learn the speakers’ accent and tone (user profile, unlimited)
  • Automatic speaker recognition ability to select user profile automatically (Avoice service)
  • Persian text-to-speech ability to read the typed text (Atalk service)
  • Environmental adaptation capability to maintain performance in noisy conditions
  • Supports voice commands to run and control applications (Atask service)
  • Providing record history for upcoming edition by physician or secretary
  • Writing numbers and punctuations such as points and question marks by talking
  • A customized editor with special facilities such as Atext service
  • Coming with a special speech recognition headset
  • Bilingual user interface (English and Persian)
  • Having single-user and multi-user network versions

 

Atype-Legal Package includes:

  • Software CD
  • Atype user guide CD
  • Speech recognition microphone
  • USB lock
  • User manual booklet

 

Atype-Legal requirements

  • Windows XP operating system or higher
  • Installation of Persian Language and keyboard
  • Dot Net 3.0 Environment (to be installed along with the software)
  • 2 GHz CPU or higher
  • 512 MB of RAM for Windows XP and 1 GB for Windows Vista and Win7
  • 250 MB of free disk space
  • Sound card with microphone input
  • High quality speech recognition microphone (to be supplied with software)