Bertin IT has launched an upgraded version of its MediaSpeech transcription platform. In addition to being faster and more accurate than its predecessor, MediaSpeech v6 is compatible with 17 different languages, allowing users to create searchable (and analyzable) text transcripts of video and audio content.
The updated version of the platform needs only milliseconds to transcribe each word, allowing MediaSpeech v6 to deliver real-time transcription of live events. It also includes gender recognition and can reference a biometric database to identify individual speakers.
MediaSpeech is available through the cloud, on site, or as a software service, with a Factory version for customers who need to process a large volume of content. Meanwhile, the platform’s real-time performance capabilities allow users to leverage MediaSpeech for chatbots and call centers.
Of course, MediaSpeech is one of several voice-to-text solutions currently on the market. Red Box recently integrated Nuance’s Transcription Engine into its conversational capture platform, while Microsoft’s speech recognition engine has achieved human parity on more than one occasion. Google has made similar strides, releasing speech recognition software that outputs every individual character and is small enough to fit onto a mobile device.
That’s not necessarily a bad thing given Grand View’s predictions for a robust speech and voice recognition market. If the research firm’s $31 billion figure proves to be accurate, there should be plenty of space for smaller companies like Bertin IT.