DikteAPI - based on KTM® (registered Speech Recognition Engine of Yöndata) - is a software component prepared for software developers. It is entirely developed in Turkey and for Turkish language. By using DikteAPI, you can perform

• Reporting
  
• Menu Navigation
  
• Command Control
  
• Data Entry
  
• Program Management
  

and many other activities in your program via the user's voice.

With:

• Native Delphi Component
  
• ActiveX Component (OCX)
  
• PowerBuilder User Object
  
• Dynamic Link Library (dll)
  

alternatives, DikteAPI can easily be integrated with a wide variety of software development platforms.

DikteAPI is available in three different packages; DikteAPI Light, DikteAPI Standard and DikteAPI Professional, accordingly to the fields of use, with various alternatives.

• DikteAPI Light:

DikteAPI Light is developed using Isolated Speech Recognition and allows navigation through menus, recording of statements and of controlling command functions using the users voice.

Because the system is developed on the basis of word recognition, in order for the sentences to be recognized, stopping between words are a necessity.

DikteAPI Light is the most accurate version in speech recognition and is the most user friendly considering command control

For example ,it can easily differentiate words even if there is a one letter variation among them, such as it is in the case of "dikteleştirmedik" and "dikteleştiremedik".

You can dynamically establish and input as words the commands like “bilgisayarı kapat”, “ekranı kapat”, “epostayı aç”, “eposta gönder”, “dosya aç”, “mizan ver”, “personel devam çizelgesi” and eventually make it recognized by the program.

Note: Suggested maximum number of words is 500 as the performance will incline due to the rise in the words being input.

• DikteAPI Standard:

It is a version of DikteAPI Professional with word capacity limited to 1000.

If desired, DikteAPI Light included in the package can also be dynamically used.

DikteAPI Professional:

DikteAPI Professional is developed using Very Large Vocabulary Continuous Speech Recognition, transforms the words being spoken by the user to written form in real time without necessitating pauses between words even if the speech is given in everyday language using thousands of words.

Let alone the fact that the number of words doesn't result in lower performance, but also the program is continuously getting more accurate due to the acoustic adaptation and grammar education. Among all versions, DikteAPI Professional is the most untroubled considering high speed data entry without using the keyboard or the Mouse.

If desired, DikteAPI Light included in the package can also be dynamically used.

Note: Words been altered by adjuncts are considered independently, as new words.

Technical Specifications:

• Recognition / process capacity of 65,000 words with its unique Very Large Vocabulary Continuous Speech Recognition technology
  
• Up to 95% of accuracy rates for sentences spoken in daily conversation speed and structure
  
• Over 95% of accuracy rates for individual words or word groups
  
• High recognition rates
  
• Artificial Intelligence and Learning System infrastructure
  
• Dynamic vocabulary management
  
• Easy integration with software development process
  
• Rapid application development advantage with less number of functions / methods
  
• Full compatibility with Microsoft Windows operating system
  
• Easy usage with all well known application development tools
  

System Requirements:

• A computer with 2.4 GHz - Hyper Threat processor or more
  
• 512 MB RAM or more
  
• Minimum 200 MB of available hard-disk space
  
• Microsoft Windows 2000, XP, Vista operating system