VOice REcognition RObust (VORERO) (267)
Speech Recognition Middleware
- Speaker-independent speech recognition middleware
- Compact HMM base architecture
- Noise robustness
- Available for majority of 32-bit MPUs
- Languages: U.S. English, Japanese, Korean, Mandarin
VOice REcognition RObust (VORERO) is speaker-independent speech recognition middleware. It provides compact size, small memory footprint, and excellent noise robustness so that it is suitable for embedded speech systems such as telematics, navigation, PDAs, and home appliances. It supports a series of 32-bit RISC processors, including ARM and StrongARM, SH, MIPS, MR, and other platforms as requested.
VORERO recognizes not only a simple word, but also continuous words and continuous digits, and provides practical word spotting and out-of-vocabulary rejection features. VORERO works in quiet office or loud car environments, and can handle several words in a tiny speech system to hundreds of thousands of words for navigation scalability. It currently supports U.S. English and Japanese, and development is underway in Korean, Mandarin, and U.K. English. Support for other languages is also planned. Users can develop their own speech recognition system easily by using VORERO SDK, which includes a well-developed grammar build-up tool and optimizer tool.
For further information, contact the Voice Interface Project at http://www.vorero.com.
If you are interested in more information on how Wind River partners with this company, click here.