SpeechTech MegaWord is an application for automatic transcription of spoken dictation into written form. We use our own highly advanced technology for automatic recognition of continuous speech, which can be integrated into other products. High quality product in the transcribed dictation MegaWord greatly help quality domain specific language models. For example, in "judicial domain" system achieves 97% success rate. On the other domains are working in cooperation with our partners and customers.
MegaWord system uses the latest results of a research project MegaWord.cz, which takes place in cooperation with the University of West Bohemia. (Project leader is SpeechTech, Ltd., Deputy Director is the University of West Bohemia). The aim and objectives of the research project MegaWord.cz is research and experimental development of a new generation of software for automatic conversion of speech into text (or "dictating system"). It is a system with a wide experience which allows acceleration and facilitation of training -" Production " - text documents. The system will operate in real time with the dictionary significantly more than million words obviously providing the necessary user-friendly interface. This project was funded with support from the state budget through the Ministry of Trade and Industry.
One of the branded version of the product MegaWord is the NovaVoice® software, which is distributed by Consulting Company Novasoft a.s.. This product has been awarded by Crystal Drive in the exhibition Invex 2008. Then in 2009 the company acquired the product Czech Head award.
Advantages of automatic transcription of spoken texts with dictation recording the largest in the transcription complex texts, which will reduce administrative elaborateness and accelerate the handling of documents, laboratory or other reports and resolutions, etc. MegaWord in this regard offers the judicial domain, and even specialized domain terminology of regional offices.
Using the product MegaWord we develop specialized solutions for healthcare - working with leading clinical practice for diagnostic work in the fields of pathology (autopsy reports, biopsy, necropsy), radiology (X-ray), surgery and others. Language models include solutions for voice control system, dictation abbreviations, codes, inserting pre-text portions of standard voice command. The resulting text form does not contain typing errors and recognized words are formulated under the normal rules of medical practice. Here is a screenshot of the program where you can find an example of recognized academic text.
MegaWord can be adapted for use in other terminological domains where there is a the text resources for creating specific language models. The system is developed as independent on speaker, but if necessary the acoustic models can also be adjusted to a specific speaker, which can for example serve better to people with tiny speech impediments or unusual pronunciation.
To purchase a system MegaWord please contact us.