Technology

Speech enhancement technology can be applied always if our communication is affected by environmental or other external noise sources. Speech enhancement is a long standing research topic engaging leading scientific institutes and commercial companies. However, such systems are steadily improved. Due to our long lasting research in speech enhancement we can offer most advanced and efficient solutions.

Our NOISEBUSTERS technology is based on advanced statistical signal processing methods. Distinct feature of our algorithms is a build in mechanism of automatic adapting the noise profile. The method adapts to  changing with time noise characteristics eliminating the need for user intervention. Additionally a method is implemented, which equalizes the recording level.
We also work on experimental speech enhancement methods, which use the so called a-priori knowledge about the speech signal.

We have samples of enhanced speech using our technology. We are ready for a realization of custom, individually tailored speech enhancement systems.

Automatic speech recognition (ASR) technology can be used for convenient human man-machine interaction by voice commands or even continuous speech. Due to ASR we can control devices, have access to databases or dictate SMS. The ASR systems handle the task of changing spoken word into text. Building of such system is technically more complex than building a text to speech system. There are several means to accomplish the goals of ASR.

The VOICE LAB company uses statistical analysis of the speech to be recognized with help of the Hidden Markov Models (HMM). We use open-source as well as our own solutions -dependent on demands and tasks. We specialize in collecting databases of speech material, and deploy ASR systems as on-line e-services. We are developing also methods for global optimization of Hidden Markov Models and deploy such systems on high performance computing platforms.

Based on our experience and solutions we are ready to deploy ASR systems in the following areas:
• for mobile application with the Android, Symbian and iPhone
• for call center applications
• for navigating web-pages
• for other specified, on demand applications