Tag Archives: Speech recognition

Hospitals Lead In Speech Recognition Infrastructures

Posted on by

If you consider all the fields of human activities where speech recognition technology might be used, the hospital environment would seem to offer the most beneficial uses.  Just think of some of the benefits:

  • Reducing cross-infection by using keyboards
  • Hands-free operation allowing the physician to minister to the patient
  • Reduction in errors in record keeping for patients
  • Improved quality of patient care
  • Reduced cost

Not surprisingly, it does not seem inappropriate to say that Every hospital will have speech recognition infrastructure.

A big influence on that will be Nuance  It is among the world’s most influential software companies with 4.5 billion users – half the world’s population. It specialises in data input and information capture by text or speech – improving communication between man and machine. Following its recent acquisitions Nuance have now embarked on their mission across the European healthcare systems.

The Nuance vision of how speech recognition will be used in five years’ time is that it will change the way health services are delivered in a major way:

In five or ten years, all hospitals will use electronic patient record systems. These will become more structured so we can get statistical information from them easily. At the moment, we can’t, because the data is sitting on paper and is virtually unusable.

Speech recognition will continue to be a major help to doctors filling in these structured reports.  We are already researching ‘talk forms’ – doctors talk in free narrative and the form fills in automatically.  We also have to move towards using decision support systems to give immediate feedback to doctors when they make a mistake – at the point of dictation.  Whether that happens over an iPod, a PDA or a mobile phone, it won’t matter – we will be there because in the future, every hospital will have speech recognition infrastructure.

Just check out the technology being used the next time you enter a hospital environment.  You may be surprised what you hear.

Enhanced by Zemanta

Technorati Tags: , , , ,

Do You Hear How Google Will Search?

Posted on by

Virginia Nussey had an interesting interview with Matt Cutts on The Future of Search. Matt Cutts is the head of Google’s Webspam Team and an all-around authority on Google guidelines and practices.

He had a deal to say about spam but the most intriguing section was on Mobile Search.

2009 will be a big year for mobile.

Yes, another big year for mobile — but for real this time! It’s not that the last several big years for mobile were hoaxes, but rather that we’ve been building to this point for a while now, and the time is finally upon us. The Pew Internet Project, an initiative of the Pew Research Center, released The Future of the Internet III report this week and the verdict is in.

By 2020 (and maybe even before that), mobile devices will be the primary connection tool to the Internet. This is a prediction that has rightly scared a few people who realize that this could blur the line between work time and personal time even further. Matt’s concerns lie elsewhere though, focusing on how to make search useful on a small screen and on things like the progress of speech to text, machine translation and face recognition technologies.

Speech technology is the natural way to interact with a small cell phone.  For many, the use of keyboards with incredibly small keys is at the least awkward and in some cases impossible.  There are many challenges in using the rich signals that come from a voice but it must be the eventual winner.  Matt Cutts alluded to this, and I am certain this will prove to be the eventual solution.

Reblog this post [with Zemanta]

Technorati Tags: , , ,

Voices mumble, fingers fumble

Posted on by

John Markoff in his New York Times blog post, Google, iPhone and the Future of Machines That Listen, raises some important issues. He is talking about Google’s new speech recognition service for the iPhone, which was released on Monday. 

He suggests that it will understand you most accurately when you speak to it just the way you enter queries into the Google search box. .. The accuracy is far from 100 percent, and probably not even 95 percent (Google execs demurred when I asked if they had any meaningful accuracy statistics). My experience is that it captures your voice query substantially more than half the time, and that in itself is a revelation.

More and more people seem to be finding that voice recognition systems can often do the job.  As cell phones become smaller, it seems likely that speech technology will become the preferred way to input instructions and questions.

Another important influence here may be changing population age structure.  Senior citizens may prefer to talk to their cell phone, rather than trying to hit those incredibly tiny buttons.  Even if someone mumbles, what they say includes extremely rich data.  With the right software, the cell phone can undoubtedly figure out what is wanted.  At worst it can ask a question or questions for greater precision.

On the other hand (no pun intended), if you hit the wrong key, the cell phone may still assume that this was intentional.  Trying to figure out when the cell phone should question you on whether you intended to hit that key would not be easy.  It would also lead to much frustration as the cell phone refused to accept your command.

There are challenges either way, but I believe those facing speech technology will prove to be solvable in very user friendly ways.  With all the resources the big guns are applying here, it is likely that acceptable, commercial solutions will be seen within 18 months

‘

Reblog this post [with Zemanta]

Technorati Tags: , , , ,

Search the Internet for other related articles.
Loading