Sensory's Voice Biometric Technology Breaks One Million Installed Base; ``Killer-App'' Considered to be Personal Biometric Smart Cards

    SANTA CLARA, Calif.--December 4, 2001--Sensory, Inc., the leader in embedded speech technologies, today announced the one-millionth deployment of its speaker verification system. The breakthrough volume deployment of this biometric technology is considered unique because of its ability to deliver a human biometric at a cost-point low enough to embed as an inexpensive standalone biometric system.

    The Voice Biometric Market

    Voice biometrics rank no. 1 for value proposition among DNA, Iris, Retina, Finger, Face, Signature, and Voice technologies according to a 1/22/01 report by Morgan Keegan industry analyst Brian Ruttenbur. According to Ruttenbur, "Voice is the most convenient and cost-effective approach for biometrics, and Sensory's overwhelming success is the best proof of the value added from an embedded biometric solution."
    In a March 2001 study of the Voice Biometric market, Frost and Sullivan broke the voice biometric market into two basic categories -- Telephony and Non-Telephony. Telephony approaches use a central server based verifier, and entail very heavy requirements for MIPS and memory. Sensory has been focused on the non-telephony market segment, which tends to be a PC based or embedded approach to verification. Frost & Sullivan expects the non-telephony market to grow by a 58.84% compounded annual growth rate from 2000-2007, and says today's market is dominated by Sensory, with a market share of greater than 30%.
    A recent market study by J. Markowitz and Associates predicted the voice biometrics market to grow to $300 million by 2005.

    The Low Cost Sensory Biometric System

    The Sensory speaker verification technology can be implemented in low cost chips or as a software add-on to any existing DSP or microcontroller-based system. Sensory's text-dependent speaker verification technology (voice password) requires fewer than 10 MIPS of power and as little as 20KB of memory. Unlike other biometric solutions requiring expensive scanners or 3-D readers to encode the biometric information, the Sensory solution can be implemented with a less than $0.10 microphone element, making the total implementation cost in high volume applications using a Sensory IC available for under $2.
    "Sensory's technology is so small and so inexpensive that it can actually be implemented in a stand-alone biometric reading smart card. Having a fully self contained biometric device like a personal smart card offers substantial security advantages over devices that must hook up or connect on-line; the standalone device essentially can't be broken into by hackers, making identity theft virtually impossible," says Todd Mozer, Sensory's CEO
    Sensory also has an award winning speaker identification technology, which can listen to a person talking and identify the right person from the characteristics of their speech. Both speaker verification and speaker identification require some degree of user training. The verifier uses a simple word or phrase based training, and when verifying is checking to see if an individuals identity is valid, while the identifier trains on continuous speech data, and is able to identify one person out of many.

    Applications

    Sensory's biometric technologies can be used in a variety of embedded and consumer applications for security, including automotive, home, and personal electronics (for data access). Recent implementations of Sensory's biometric technology have included an ignition system that won't start unless the right person is present, a telephone with voice-locking features, a girl's diary that can only be opened by the right user, and a security alarm that must be turned off with a particular persons voice.
    According to Judith Markowitz, industry analyst and editor of Voice ID Quarterly, "The biometric smart card, with its embedded speaker-authentication and self-contained microphone is a creative approach to portable, user-controlled security. It fits well into Sensory's existing line of low-cost, embedded technology and enables the lowest cost method of creating a biometric smart card."

    About Sensory, Inc.

    Based in Santa Clara, Calif., Sensory, Inc., is the leading provider of high-quality, low-cost speech-recognition and speech-synthesis technology for embedded applications. Sensory's speech technology is used in personal electronics, Internet appliances, interactive toys, wireless and automotive applications. Sensory's Interactive Speech(TM) line of products includes the award winning RSC series speech chips, small-footprint Voice Activation software, Fluent Speech(TM) large-vocabulary software engine and Fluent Animated Speech(TM) technology. Sensory's customers include leading companies in the consumer-electronics and embedded-product markets, such as JVC, Hasbro, Kenwood, Mattel, MGA, Radica, Sega, Sharper Image, Fisher-Price, Sony, Tektronix, Toshiba, Uniden, and VOS. More information is available from Sensory's web site at www.sensoryinc.com.