Wednesday, July 28, 2010

Shout at your computer and it will listen



The days of the physical keyboard are numbered. This ancient device is being replaced by digital keyboards, rollup keyboards thin as a piece of paper and speech recognition technology. The future of typing is not better keyboards - it is no keyboards at all.

In the last five years, voice recognition technology has come a long way and entered the mainstream. It is now common in medical and legal fields and is available on most computers and the latest smartphones. Windows Vista and 7 have Windows Speech Recognition, which allows the user to dictate and control the computer through voice commands. The software is quite sophisticated - for instance, even when there are no obvious command buttons, the user can still control the mouse by dividing the screen into a numbered grid and speaking the coordinates of where he or she wants to click. Similarly, the Apple Mac OS X operating system also has sophisticated voice-to-text and text-to-voice software as standard.

There are also many software companies making voice control software. There are too many to mention, but one of the most popular and widely used applications is Dragon NaturallySpeaking by Nuance Communications. Although Dragon NaturallySpeaking has been around for more than twenty years, the early versions could not handle natural speech with no breaks between words but today’s versions can handle continuous, natural speech with ease. However, the software does not just transcribe audio or convert text into an audio stream; it also accepts commands from the user, allowing someone to browse the Web, for instance. With 99% accuracy and the ability to ‘type’ at more than 120 words per minute (the average is 35 words per minute), the software can save you a lot of time. But it’s not just for lazy or busy people; many disabled people use such software, which is also popular in the medical industry for transcription.

Cellphones are also forging ahead with speech recognition software - the iPhone is one of the best examples. It features voice-activated dialling and can control the iPod application via voice. Meanwhile, Microsoft has developed Microsoft Voice Command software that can control Windows Mobile devices by voice. But it is not just the big players in the smartphone market that are tapping into voice control. For example, there is Nuance Voice Control: software that can be added to any cellphone handset to allow speech driven access and control of the device. It can be used for messaging, navigation, searching the Web, playing music and so forth.

Researchers are working hard on improving the ability of cellphones to convert text to speech. In February this year, researchers in the United States created an application called VoiceTEXT that lets people speak text messages instead of typing them. The message is then converted to audio or transcribed and sent as an SMS, depending on the abilities of the receiving phone. While the software is not revolutionary, it can be used on any phone and, experts say, will help create safer roads by letting people concentrate more on driving than typing out messages.

Speech recognition software has become mainstream thanks to Microsoft bundling the software as standard with Vista and Windows 7. It has also crept onto many other computerised devices, from landline phones to cars (Fords in particular). Now we just have to wait for keyboards to disappear completely.

No comments:

Post a Comment