Purpose of the development: To simplify the process of converting sound to text.

Recommended application field: For stenographers, journalists and students.

Technical characteristic: Signal to noise ratio - more than 20 dB. Recognition performance for one core - less than 0.5 sec. for processing 1 sec. speech. The volume of the dictionary - 200 thousand words.

Advantages over analogues: Distributed automatic recognition of spontaneous speech as a web service. Record sound from a microphone to files. Convert videos to text with You Nube. The word-for-word accuracy of bagato-speaker recognition of spontaneous speech is more than 80%. The ability to adapt to the voice of unfamiliar speakers.

The development stage readiness: Tested experimental exploitation order

Description of the development:
()
Automatic speech recognition does most of the hard work of listening to sound recordings. A person needs to correct errors and form the final form of the document.

corresponds technical description
Guarantees stable results getting

Possibility of transfer abroad:
Combinated reduction to industrial level
Joint production, sale, exploitation

Photo

Country Ukraine

For additional information turn to:
E-mail: gal@uintei.kiev.ua