Speech recognition

Dilmanc Automatic Speech Recognition System is first such system for Azerbaijani language and enables to enter the text via dictating to microphone. Input speed is 3-4 times faster than typing in a keyboard. Another feature of system is enabling voice-control of computer and other devices. It recognizes near 1 million word-forms, with accuracy of 92% . The system will be released at the end of 2012, when the accuracy will increase further.

Main Features:

• Input text via dictating to microphone
• Input speed 3-4 times faster than typing
• Recognizes near 1 million word-forms
• 92% accuracy

Applying of the system:

Dilmanc speech recognition system is one of the main components of Phone-translator which was developed within the project
»read more

The voice-controlled mini computer for visually impaired people was developed on the basis of the speech recognition system
»read more

Presentation of the system:

Dilmanc speech recognition system was firstly presented as part of the Phone-Translator system in ITU Telecom World 2009
»read more

More improved version of the system was presented in BakuTel 2012 - Azerbaijan International Telecommunications and Information Technologies Exhibition
»read more

The speech recognition system will be released in the near future.