The 2015 State of the Speech Technology Industry: Assistive Technology
have been researching this technology are not pushing it out commercially since it doesn't impact them personally. Most of the research now is coming from universities. Universities are pushing this technology further, and without their research, we would not have been able to develop this product."
CereProc, a 10-year-old Edinburgh, Scotland–based company, gained acclaim when it gave famed film critic Roger Ebert his voice back. (Ebert passed away in 2013.) CereProc was formed by two speech synthesis experts, Matthew Aylett and Nick Wright, who received research funding to create artificial voices that sounded more human than robotic and which later included emotion.
Since then, CereProc has used its solutions for a variety of situations, such as speech synthesis for humanoid robots, working with video gaming companies to use its speech-to-text solutions for visually impaired gamers, and other endeavors.
The company now allows individual users to access its speech-to-text technology via its online voice cloning solution. Using a Web browser, the program takes an individual's spoken prompts from a CereProc script. This gathered data is then deployed to create a Windows TTS version of that person's voice. CereProc said that the method of creating good results is costly, but the company's solution makes it affordable to its customers, including the disabled.
Getting an Assist from the Big Players
While some might feel that university-backed projects have invested more effort in enabling the disabled, consider that often a product that begins as an assistive solution can later be remarketed to the general public. Large entities usually have deeper pockets for research and development and can scale at a faster and broader level.
In 2014, Nuance added new capabilities to its popular Dragon products, such as a transcription feature for Dragon Dictate and a built-in microphone for Dragon NaturallySpeaking. The company's speech and voice recognition efforts have expanded to mobile technologies as well. The Dragon mobile assistant application now has a combination of voice biometrics and "a one-shot wake-up word" from a specific person, which enables hands-free interactions with Android phones, says Rick Brown, senior director of product management for Dragon at Nuance.
"A person could issue a custom wake-up word and command to a device with voice biometrics, such as, 'Hey, Jarvis, what will the weather be like tomorrow?' ensuring that only that person's voice is recognized and responded to," Brown wrote in an email to Speech Technology magazine.
It's not always professionals who are driving development for assistive speech-enabled solutions. Brown pointed out that several mobile applications for accessibility have come from Nuance's NDEV mobile developer program. One such developer, 12-year-old Eric Zeiberg, has created a handwriting-to-speech application, HandySpeech for iOS. The solution enables people to use their fingers to draw on the screens of their smartphones and tablets, which then enables text to be voiced.
Another Nuance product, TALK, is geared to the blind or visually impaired. The solution operates on mobile phones and, using text-to-speech, reads aloud text from messages, contact and caller IDs, as well as content from other sources, such as mobile browsers and maps. Users can choose either an ETI-Eloquence or Vocalizer text-to-speech voice. ETI-Eloquence comes from Nuance's partner the Code Factory, and is an Android version of a TTS synthesizer. It includes support for 10 languages and offers GPS solutions such as Google Maps, as well as translators and e-book readers.
At Intel Labs, renowned physicist and cosmologist Stephen Hawking relies on a software system that was recently upgraded by the company's scientists. While the resulting improvement in his communication system primarily involved updating his computer interface, Hawking still uses a voice synthesizer from Speech Plus, a company that is no longer in business. Even so, Hawking and Intel reiterated the message that newer technology is greatly helping to close the gap between physically challenged people and the able-bodied.
Related Articles
In our first State of the Speech Technology Industry issue, we reveal the latest trends and developments in eight market categories.
10 Feb 2015