Technology
Virage SoftSound technology was originally developed at Cambridge University. Virage SoftSound
licensed, commercialized and further developed this technology to provide a broad base of audio
and speech processing algorithms. The key features of our speech recognition technology are:
- Speaker independent operation - works out of the box without manual training
- Very large vocabularies - no arbitrary limit on size
- Patented search technology to drastically reduce CPU and memory requirements
- Automatic customisation via additional text material
- Fixed-latency real-time mode for media monitoring operations
In addition we have developed technology for:
- Speaker change and speaker recognition
- Automatic synchronisation of text and audio for subtitling and talking-book applications
- Word spotting and phrase recognition
