Mobile Audio Data Recording

The functionality of speech recognizers is mainly determined by current noise environment. The worse the signal-to-noise ratio, the more difficult it is for speech recognition to work effectively. There is a huge difference between the signal-to-noise ratios when using speech recognition with a headset microphone in a quiet office or using speech recognition without a headset microphone in a busy train station. Therefore the speech recognizer must be trained to cope with different environments.

For this reason it is necessary to train the speech recognizer using speech data recorded in the desired environments.

voice INTER connect has developed a mobile audio lab which makes it possible to record speech data in different environments. Besides this it is possible to test several microphones on their suitability for speech recognition. Furthermore, speech data can be recorded simultaneously with different microphone arrangements.

Features

• Easy transportability, high mobility (recordings can be made in cars, public places, offices, etc.)
• Operates with 12V direct voltage or 230V AC voltage
• Rapid startup time
• Intelligent cable layout
• Simultaneous speech recording with up to 8 microphones
• Use of high quality microphones with different directional characteristics
• Integrated software functionality for rapid cutting, labelling and converting of recorded speech data into the desired format
• Recordings with up to 48 kHz sampling rate and 24 Bit dynamics are possible