Speech in Consumer Products

With the introduction of speech-enabled cars, toys, cellular phones and other electronic devices, speech technologies having been gaining recognition in consumer products. There are several products on the market today that are affordable and becoming more commonly used. Whether appealing to children with toys and dolls (see sidebar, “Playing with Toys”) or adults with automobiles and cell phones, the companies that use speech are beginning to see more and more usage by the average consumer. With services such as the well-known cellular phone offerings by Sprint and most other major wireless carriers, voice activated dialing (VAD) has been available and usable for some time. Most carriers offer VAD in which the recognition work is carried out in the network, rather than inside the device itself. The amount of contacts that could be stored in the network are virtually limitless, and the processing power is easier to manage than offering it on the device. However, this may be changing and appears to be following Moore’s Law. As processing power has increased on smaller devices, so has that of speech recognition. Embedded speech now allows the devices to hold larger contact lists and are speaker-independent, allowing any user to easily navigate a device with little or no training. The A500 cellular phone, offered by Samsung with speech technology powered by Voice Signal Technologies, allows voice dialing by speaking the stored contact name or by spoken digit dialing. The phone also allows for speech-driven features such as voice mail retrieval and Web browser initiation. The product retails for about $299 from Sprint. ART Advanced Recognition Technologies Inc. has several customer wins in the cellular phone market, all of which use ART’s smARTspeakXG speech recognition product. All of the speech-enabled phone offerings by manufacturers such as Qualcomm, LG, NEC and several others utilize ART’s speaker independent name dialing solution. And according to Eran Aharonson, ART's president, these phones can have voice-enabled contact lists in upwards of 200 names. In some Smart Phone and PDA/phone combination offerings contact lists can reach into the thousands of entries. This combats one benefit that network-based recognition systems had over embedded systems in the past. Speech technology-enhanced products have also been marketed to the homeowner. At stores such as Home Depot and Orchard Supply Hardware, consumers can purchase a voice-activated dimmer switch that takes the place of the normal wall switch for controlling room lighting. Offered by Vos Systems and speech-enabled by Sensory Inc., it allows both speaker independent and speaker dependent operation giving the user some flexibility. There are five speaker independent programs to choose from, with the speaker dependent operation allowing the user to program other commands or languages. The product retails for about $50. Vos Systems also offers other dimming and on/off functional devices for household lamps in approximately the same price range as the wall unit. From retailers such as Brookstone and Hammacher Schlemmer, a voice-enabled remote control is available from a company called InVoca. With speaker-dependent technology powered by Sensory, the 8-in-1 remote can control your TV, VCR, CD Player, etc., with programmable voice commands. The device can store up to 50 commands and recognize the voices of four people. The remote’s functions are programmable for storing multiple functions into one voice command. For example, with one command the TV can turn on, be tuned to channel three, while the VCR powers on and begins to record. The pricing of the remote is variable depending upon the retailer, but ranges between $50 and $70. Automobile manufacturers have also been seeing the benefits of speech technology. Infiniti and Jaguar have been using speech recognition for the eyes-free/hands-free control of such things as climate and music-function control. Voice recognition features in the Infiniti line are available with the M45 and Q45 models and in the S-Type of Jaguar. The price ranges for these automobiles vary depending upon the product line and options chosen. Both of these cars have speech recognition technology powered by SpeechWorks in conjunction with Visteon Corporation. At the recent North American International Auto Show 2003 held in Detroit, Ford unveiled a concept car that could be taking speech recognition to a new level for in-vehicle control and interaction. The Model U Concept Sport Utility Vehicle, with speech technology powered by SpeechWorks, demonstrated some unique new functions and ideas for vehicles of the future. With this new concept car Ford is focusing on a more “conversational” approach to speech recognition in the vehicle. Where predecessors such as Jaguar and Infiniti allowed for a limited set of vehicle features to be controllable by speech, this new SUV allows the driver to interact with it in a more dialog-oriented fashion. System functions that can be controlled include navigation, cellular phone (any Bluetooth-enabled phone), entertainment (i.e. cd player, radio, etc.), climate control, the retractable roof, and personalization preferences for additional drivers. The TTS system was also designed to be more human-like with hopes of creating a more conversational user interface. Developers feel that a major improvement over previous speech-enabled systems lies in the new navigation capabilities. In contrast to most systems that require typing in a location or spelling it out letter-by-letter, number-by-number, you can tell the Model U a specific address, major intersection or one of many points of interest, and it will audibly guide you to your destination. Ford sees the Model U as having the potential for revolutionizing personal transportation, in possibly the same way the Model T did nearly 100 years ago. (For more information on speech technology in automobiles, see “Speech and the Automobile” in the November/December 2002 issue of Speech Technology Magazine) Voice activated home appliances, toys, automobiles, etc. are increasingly finding their way into the homes of average consumers. According to The Kelsey Group in its “Embedded Speech Technologies and Overview” report, licensing revenues of embedded speech technologies will have grown from $8.3 million in 2002 to $276 million by 2006. With this kind of expected growth, one can expect to see more and more households and automobiles using speech technologies. It will be interesting to see what consumer product manufacturers will introduce using speech technologies over the next few years, as well as which products will be enhanced with speech.
One Retailer’s View…
"The Sharper Image has considered integrating speech technology into nearly every kind of product one could imagine. “It is our experience that [speech] technology fascinates consumers, but the products of the late 80s and 90s were frustrating to use for most people. With some of the low cost solutions available today, toys may be the biggest opportunity for voice-activated products. Right now we see the chips and software are getting substantially better - this should help consumers and manufacturers to gain confidence with voice technology. These two trends mean that in the near future it will be easier to create voice technology products more quickly and inexpensively than ever before, and the consumer will be eager to try (and be satisfied with) the result.”
- Andrew Parker, vice president of Engineering & Product Development, The Shaper Image The Sharper Image is a specialty retailer that is known for being a source of new and innovative consumer products.
Mike Terry is the editor of Speech Technology Magazine. He can be reached at mike@amcommpublications.com
SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues