NovoSpeech Receives a Patent for Its Quantum Search Technology
NovoSpeech, a portfolio company of The Trendlines Group, received a notice of allowance from the U.S. Patent and Trademark Office for its quantum search patent, an integral component of the company's speech recognition technology for the handheld device market.
NovoSpeech's quantum search application solves the problem of memory and computational resources on handheld devices. In large-vocabulary speech recognition implementations, converting spoken utterances to text requires a great deal of memory and computational resources. This does not pose a problem on a computer or on a server, but on handheld devices, both memory and computational resources are far more limited.
NovoSpeech has adapted a quantum computing search algorithm specifically for speech recognition searches. The company's algorithm reduces the complexity of the search in the linguistic model, facilitating searches through very large vocabulary speech recognition implementations, even on handheld devices.
"The accelerated manner in which all of our patent claims were deemed innovative, inventive, and industrially applicable shows that we are pioneers in the quest to provide very large-vocabulary speech recognition that runs on platforms with memory/computation constraints," said Tina Ornstein, CEO of NovoSpeech, in a statement. "We are currently seeking additional strategic partners to share in the development of our next-generation product, a quantum speech recognition decoder that will revolutionize embedded speech recognition."
"This is another significant milestone for NovoSpeech and furthers their mission to transform speech recognition. As the mobile device becomes an integral part of our lives, the ability to provide quality automatic speech recognition in different surroundings, across languages and accents, will become a competitive factor for anyone in this market," said Steve Rhodes, chairman and CEO of Trendlines, in a statement.
An Israeli company, NovoSpeech is dedicated to developing a speech recognition engine that provides highly robust, large-vocabulary speech recognition entirely on-device without the need to connect to the cloud.