Press Release

TravellingWave Solves Multimodal Noise Interference Issues; Introduces Innovative Speech Processing Algorithm for Speech-to-Text

TravellingWave, a leader in next-generation mobile user-interface technology, today announced that it has solved multimodal noise interference issues commonly associated with mobile speech and text input through the development of a new speech processing algorithm.

Source: TravellingWave
6/25/2008

Noted Wireless Executive Todd Achilles, Formerly with T-Mobile and HTC-Americas, Joins Advisory Board to Spear-head Delivery of VoicePredict™ in Second Half of 2008

SEATTLE, WA – June 25, 2008 – TravellingWave, a leader in next-generation mobile user-interface technology, today announced that it has solved multimodal noise interference issues commonly associated with mobile speech and text input through the development of a new speech processing algorithm. The company also announced that Todd Achilles, former Executive Director of Handsets for T-Mobile-USA and former General Manager for HTC-Americas, has joined TravellingWave’s advisory board and will assist management in product innovation and marketing strategies. TravellingWave is currently in trials with multiple mobile phone manufacturers and mobile carriers for VoicePredict™ which will be additionally released in the second half of the year as a downloadable software package for end-users with selective handsets.

TravellingWave’s Rao-Aronov-Garafutdinov speech-processing (RAGs) algorithm enables the VoicePredict product to operate in a multitude of noisy environments. RAGs is an innovative speech processing approach that extracts frequency localized time-varying modulation features, which reduce background/channel noises; mitigate noisy interferences (as in low frequency hums); and suppress impulsive noises (for instance, clicks that are picked by a mobile microphone when users tap the keys). The algorithm results in robust speech recognition accuracies, creating a more user-friendly solution for mobile consumers, who can more readily speak and type on their mobile devices - even in crowded, noisy environments.

“We have engineered VoicePredict to become the industry’s first next-generation multimodal mobile user experience,” said Ashwin Rao, founder and CEO of TravellingWave. “We are bridging the interface gap between human-hand and human-voice interaction.”

At one end of the spectrum, traditional speech-to-text technology has been plagued by numerous challenges including background/channel noises and variability in users’ speaking styles/levels. These challenges have yielded poor accuracies that have, for the most part, resulted in an inconsistent user experience. At the other end, predictive text technology for mobile input (generally accompanying keypad, stylus, and/or touch interfaces) continues to lack the speed, ease, and accuracy of a full-size keyboard. TravellingWave, using its novel framework “predictive speech-to-text™”, connects these two technologies to present an enhanced multimodal mobile user interface: VoicePredict.

Using VoicePredict (a software that resides on mobile devices), a user simply speaks a word, types a letter or two, and the word appears on the mobile screen; if the user opts not to speak but only type, then the word still appears on the screen via standard text-prediction, which requires relatively more typed letters. Additionally, VoicePredict provides users a choice to simply speak in order to enter symbols, emoticons, contact names and standard phrases.

RAGs algorithm is based on the founding principle of TravellingWave -- the “Travellingwave phenomenon” -- in the human auditory system, well-known to scientists in the field of cochlear mechanics. RAGs involves advances in areas of time-frequency signal processing, computational mathematics, computer science, and statistical speech recognition.

Since its August 2007 industry launch, TravellingWave has announced an “Always Listening™” mode that has taken VoicePredict one step closer to a seamless multimodal mobile user experience. Under a continuing series of grant awards from the National Science Foundation, the company has furthered its research to address speech recognition robustness in a range of real-world noisy environments.

“I have seen many approaches to solving the mobile speech-to-text challenge, all of which have been frustratingly inconsistent. The TravellingWave approach flips the problem on its head and provides a major leap forward in the user experience for mobile consumers,” said Todd Achilles, Managing Director for Affinity Mobile and an industry leader on mobile applications and phone design.

VoicePredict was named Washington State’s “Breakthrough Technology of the Year” in February by the Washington Technology Industry Association – the largest statewide association of technology companies, IT departments and individual technology professionals in North America – at the Association’s 13th annual Industry Achievement Awards (IAA).

About TravellingWave: TravellingWave is an early-stage company that develops software, using speech recognition technology, for entering text into mobile devices. The predictive speech-to-text technology combines traditional predictive text input with speech recognition. The result is an extremely simple, fast, and enhanced interface for mobile users. Founded in Seattle in 2004, TravellingWave is based in Seattle, Washington. The company has been privately funded by the company founder, Dr. Ashwin Rao, and prominent Seattle-based angel investors. TravellingWave has also been partially supported by grant-awards from the National Science Foundation. For more information, please visit the company Web site at www.travellingwave.com.