As Todd Mozer, CEO of Sensory, explained to us, “version 1.0 was primarily a trigger,” meaning that its primary use involved detection of a wake-up word or phrase to activate a device or initiate a predefined interaction (like dictating an email, SMS, Tweet or Facebook posting; or initiating a mobile search). TrulyHandsFree Voice Control 2.0 adds “multiphrase” technology, meaning that it will be able to recognize and respond to a broader variety of words or phrases in order to “wake up” and respond accordingly.

The ability to distinguish spoken words from background noise is a major advancement. Rev 2.0 couples speech detection with a model for phrase recognition which enables a mobile device or toy to, in Mozer’s words, “reject the wrong words and accept the right words” in order to respond accordingly. Sensory has been working on this, more resilient “hands free” approach for a number of years. It is significantly different from Rev 1.0 of Sensory’s product as well as competing products. The additional accuracy and coincides with equally impressive advances in text-to-speech rendering and device-based speech recognition (and “understanding”). As a result, Sensory expects this approach and user acceptance to be prevalent in a two year period in cars, the kitchen and kid’s room, running on devices and products from the likes of Samsung, Mattel and others.

The technology is initially available on Android-based phones, devices and browsers. The result will be a dramatic change in user expectation regarding what end-users can accomplish without pushing any buttons.

As we noted here, Nuance has been dedicated to simplifying and streamlining the implementation of applications that provide a “natural language” interface for users. Pre-testing can prove to be an important component to such engagements because they will enable the company to detect where latencies or system delays may occur and correct them before they become a source of customer frustration.

Empirix began offering its testing-as-a-service in mid-2009. It formed a similar partnership with Avaya and its go-to-market partners in October 2010. Opus Research sees this sort of testing and ongoing monitoring as vital for promoting the highest-quality user experience across a broad spectrum of self-service and “assisted self-service” (meaning those that involve live agents) interactions.

TomTom is, by no means, the only horse in the natural language race. Nuance Communications, for instance, has VoCon “One Shot Destination Entry” baked into the nav systems in prestige nameplates, including Audi, BMW, Ford and Mercedes-Benz. In February, the company began offering a pre-packaged SDK (Software Development Kit) called VoCon Navigation, which it calls a “ready-to-integrate pre-packaged voice application module for embedded one-shot destination entry.”

As we see state troopers and local law enforcement cracking down on distracted drivers, the simplification of destination entry for navigation systems is becoming a valuable attribute.

Google, under the guidance of Mike Cohen, is going to make sure that speech recognition is one of the input options for every activity through an Android-based device or any device running a Chrome-based browser. As reported in Google’s Chrome blog, the beta version of Chrome 11 takes advantage of the W3C’s (World Wide Web Consortium’s) “Speech Input API” in order to support speech-to-text transcription which, in turn, allows for spoken navigation, dictation and command entry.

As I’ve noted in previous posts, the technology works “okay.” It is not 100% accurate and, like all speech recognition resources, it can appear to discriminate against women and individuals whose voices are in registers that correspond to a small sample of collected utterances. But speech-to-text transcription is the gateway technology into a slew of speech-enabled, mobile activities and we expect it to be ubiquitous and its use to grow geometrically as new devices, applications and use cases abound.

Both input modalities provide quick, convenient ways for holiday shoppers to comparison shop, spanning the World Wide Web and the Real World. MeMeMe, as the verbiage on its Web site indicates, is on a mission to “voice-enable any mobile application or networked service.” That makes it direct competition to Nuance, Microsoft/Tellme, AT&T/Watson, Loquendo and a few others. The Buy.com app for the iPhone is a good demonstration. It also shows how speech-based input is a peer to (but not a replacement for) keying in info or scanning barcodes.

Earlier this year I coined the term “No rest for the RESTful” to dramatize how new tools and API’s to support agile programming would accelerate application development and service delivery. As we enter 2011, the mantra appears to be “Speech is Sexy,” as evidenced not just by Apple’s “Help Wanted” posting, but by Google’s recent acquisition of Phonetic Arts and rapid-fire refinement of mobile user interfaces from Nuance, Vlingo (powered by AT&T’s Watson Engine), Google and Microsoft/Tellme.

The big difference in 2011 is that speech is getting more pervasive while, at the same time, it is being subsumed into multimodal user interfaces. Microsoft, for instance, continues to call speech recognition “foundational” to its user interface but, with the introduction of Kinect, already puts much more emphasis on accurate recognition of gestures. Last year Google’s Mike Cohen explained his objective of making speech as an alternative “every time” a keypad or keyboard is used on a mobile device.

2011 will be a year for smoothing out some of the rough spots in speech enabling the user experience. Candidates include better (more accurate) recognition, noise cancellation, more “human sounding” text-to-speech rendering, speech-to-speech translation, low-latency interaction with dynamic data “in the cloud,” and (to keep things safe, secure and personalized) voice biometrics-based authentication or ID proofing.

But there’s been a fundamental change. I used to write about the “Voice User Interface.” In the coming year attention will be on the “Mobile User Interface” that includes voice. That’s why it should not be a surprise to see that the the speech software engineer at Apple should be prepared to work on a “team” to accommodate the “rapidly changing priorities.”

AppTek brings bidirectional language pairs that include English with Arabic, Chinese, Dutch, Persian, French, German, Hebrew, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Turkish and Ukrainian, and unidirectional pairs from Indonesian, Pashto, Tagalog and Urdu into English. SAIC expects its acquired technology and assets to strengthen its efforts to deliver language services to the intelligence, defense and law enforcement communities. This complements SAIC’s portfolio of translation products that already spanned 70 languages.

SAIC is part of the Fortune 500 with over 45,000 employees. At various points in its evolution, it owned Network Solutions, which manages the DNS (Domain Name Server) that houses the majority of popular Web addresses (.com, .net, .org, .edu, for example). It has owned Bellcore and Telcordia – professional service providers that also manage the library of software and support materials that power all of the electronic switches and ancillary processors that conform to “Bell spec”.

For those keeping score: Google Voice Search exists in English, Mandarin, Japanese, Korean, French, German, Italian, Spanish, Czech, Polish, Russian and Turkish. Speech recognition is available for Android 2.2 and 2.1.

Todd demonstrates the SCIDs concept in this video:

Note that the embedded title for this YouTube video is “Truly Hands Free Trigger.wmv”. That’s because Sensory is now shipping Speech Trigger SDKs (software development kits) for mobile devices running either Apple’s iOS4 or the Android operating system. Vlingo, for one, has taken full advantage of the SDK and introduced a hands-free services called Vlingo InCar. As illustrated in the video below, the “trigger” occurs when a driver says “Hey Vlingo”. Then, in response to the prompt “What’s shakin’?” drivers can here the contents of a text message, reply, originate other messages and search for a local business to get spoken directions.

The service is optimized for the Android 2.2 operating system (“Foyo”) and uses Bluetooth for connectivity. Users of the existing Vlingo app, can also originate tweets, update status on Facebook, open and control 3rd party applications and the other activities that Vlingo supports. After a two week “beta” when the service was available only to Spring subscribers, Vlingo expanded to “all Android users” on October 19.

For those of you keeping track at home, the announcement of Vlingo InCar and the release of the “Trigger SDK” was almost simultaneous with the launch of StartTalking from
a Boston-based technology venture called AdelaVoice. As you can see in the video below, AdelaVoice laid claim to being the first and only mobile application that provides a hands-free, eyes-on-the-road way to send and receive text messages. It is a well-understood problem and one that is attracting significant investment dollars, as carriers and application providers work to support safe spoken input of text.

StartTalking Overview

Todd Mozer tells me that the same firmware that supports “trigger words” is being used in high-end, Bluetooth based earpieces from Plantronics to support a hands-free “answer or ignore” function for treating incoming calls. The phone can grab Caller ID information, render it as speech and then prompt the driver to react with “You can say ‘answer’ or ‘ignore’.” It’s brilliantly simple, and much safer than looking around and reaching down to find the phone. In both cases, the objective of the firmware is to provide the most pleasant and efficient user experience possible.