Voco

Software Voco Is a Windows application for converting speech to text. The solution Voco carries out voice typing in any text field, converts an audio file into text, includes basic and thematic dictionaries, and learns on the user's texts. The program comes out in versions: Voco.Basic, Voco.Professional and Voco.Enterprise. The main functions of Voco:

• Recognition is performed locally on the user's computer, which ensures the operability Voco Even if there is no network. Connection to the Internet is required only to activate the product and receive regular updates.
• Each version of the program Voco Includes a basic dictionary of about 85 thousand words.
• Extended versions of the program include thematic dictionaries, the number of which is constantly updated. You can also expand Voco vocabulary: for this you need to specify documents and letters - the program will learn the vocabulary and writing style used.
• Voco Professional Voco Enterprise Enable speech recognition from audio recordings. Conversion is carried out in a special plug-in for the MS Word editor with the display of audio in the player and the connection of sound with the recognized text.
• Upon command, the program will set the required punctuation mark, move the cursor to a new line or insert any text with the help of auto-replacement of the recognized command. When recognizing audio files, punctuation marks will be placed automatically.

Requirements for text editor: To operate the transcription module, which provides speech recognition from audio files, a licensed text editor MS Word should be installed on the computer in one of the following versions:

• MS Word Standard 2010 x32;
• MS Word Professional Plus 2013 x32.

Microphone requirements: The sound is input to the computer through a microphone, so the quality of the microphone and sound card affects the accuracy of recognition. High quality recognition is provided by USB headsets and microphones. Accuracy of recognition for various models of headsets:

• Jabra UC Voice 550 - 86%;
• Sennheiser Headset PC7 USB - 85%;
• Plantronics Blackwire - 84%;
• Genius I HS-02B - 74%;
• Ritmix RH-512M - 72%.

It is not recommended to use a microphone built into your laptop or webcam. Audio file requirements The accuracy of conversion of speech records into text depends on the quality of the recording device, its proximity to the speaker, the background noise environment, the correctness of speech. For spontaneous unprepared speech, characteristic for meetings and "live" communication, there are stumbling, reservations, interrupting the interlocutor or simultaneous conversation of several people, fuzzy utterance and reduction of words. Speech recognition, unlike the human brain, is not capable of "filtering" all this speech debris, and the result of recognizing such an audio recording may be low. Maximum accuracy of recognition is achieved on files recorded on Dictaphone series Gnome in dictation mode with the distance from the lips to the microphone of the recorder is not more than 50 cm. Also high recognition accuracy is achieved on the files of television and radio.