Tuesday, 21 May 2013

Best Voice Of The Week

Speech recognition has been somewhat of a UNICORN in the software industry; never really perfecting it or never able to make one without hitting a few bumps on the road. Several Pundits like Ray Kurzwell and Bill Gates have tried to develop and proclaim that speech recognition is the future of computing, but people still rely on input devices like Mouse, Keyboards, and other pointing devices.

Today, people are still curious as to the development of speech recognition software, and if they cost a few dollars or so. Several projects and companies have developed speech recognition programs over the last 2 decades, and some of them are in the 90% accuracy; meaning that even if you are able to train the software with your inflections, it still wouldn’t be good enough for others.

Every week, Linux-Tested.com will update the BEST VOICE OF THE WEEK, where we present to you the best voice recognition software in the open-source world:

Week # 1: Simon from Simon-listens.org

  • -          Simon is free open-source speech recognition software that will be able to replace your mouse and keyboard. It is very flexible software and allows customization for any other application that requires speech recognition. It is also a candidate for the European Project called E-INCLUSION because of its language-independent programming. The goal of the software is to give physically disabled people and senior citizens to do stuff on the Internet like chat, send emails, do internet banking a lot more. 

Week #2: Perlbox from Perlbox.org

  • -          Perlbox.org develops voice solution for Linux and Unix desktop control. One of their projects is called the Perlbox Voice Control with text-to-speech feature and voice control for specific applications. The text-to-speech feature wouldn’t be possible without the help of the Festival speech synthesizer. There are also desktop plugins which will allow you to control your Linux desktop with the use of your voice, including screen cycling, desktop cycling, quick-locking the screen, and invoking the run dialog. It also has a PSEUDO-COMMAND feature which allows you to enter commands that you may want to speak. There is no need to train the listening agent in this software because the Sphinx2 backend can be used by multiple users without any reconfiguration. It is also resistant to background and ambient noises. 

Linux-Tested Search

Advertisement

Banner
Joomla Templates and Joomla Extensions by JoomlaVision.Com
  • Computing

  • Latest

  • Popular

  • The Biginner's

Weather Box

Joomla Templates and Joomla Extensions by JoomlaVision.Com
  • Tue
  • Wed
  • Thu
  • Tue
  • Wed
  • Thu

Statistics

Members : 1
Content : 56
Web Links : 6
Content View Hits : 28778
Home News Popular Studies show dogs have sense of fairness

Studies show dogs have sense of fairness

What parent hasn't heard that from a child who thinks another youngster got more of something? Well, it turns out dogs can react the same way. Ask them to do a trick and they'll give it a try. For a reward, sausage say, they'll happily keep at it. But if one dog gets no reward, and then sees another get sausage for doing the same trick, just try to get the first one to do it again. Indeed, he may even turn away and refuse to look at you.

Dogs, like people and monkeys, seem to have a sense of fairness.

"Animals react to inequity," said Friederike Range of the University of Vienna, Austria, who led a team of researchers testing animals at the school's Clever Dog Lab. "To avoid stress, we should try to avoid treating them differently."

Similar responses have been seen in monkeys.

Range said she wasn't surprised at the dogs reaction, since wolves are known to cooperate with one another and appear to be sensitive to each other. Modern dogs are descended from wolves.

Next, she said, will be experiments to test how dogs and wolves work together. "Among other questions, we will investigate how differences in emotions influence cooperative abilities," she said via e-mail.

In the reward experiments reported in Tuesday's edition of Proceedings of the National Academy of Sciences, Range and colleagues experimented with dogs that understood the command "paw," to place their paw in the hand of a researcher. It's the same game as teaching a dog to "shake hands."

Those that refused at the start — and one border collie that insisted on trying to herd other dogs — were removed. That left 29 dogs to be tested in varying pairs.

The dogs sat side-by-side with an experimenter in front of them. In front of the experimenter was a divided food bowl with pieces of sausage on one side and brown bread on the other.

The dogs were asked to shake hands and each could see what reward the other received.

When one dog got a reward and the other didn't, the unrewarded animal stopped playing.

When both got a reward all was well.

One thing that did surprise the researchers was that — unlike primates — the dogs didn't seem to care whether the reward was sausage or bread.

Possibly, they suggested, the presence of a reward was so important it obscured any preference. Other possibilities, they said, are that daily training with their owners overrides a preference, or that the social condition of working next to a partner increased their motivation regardless of which reward they got.

And the dogs never rejected the food, something that primates had done when they thought the reward was unfair.

The dogs, the researchers said, "were not willing to pay a cost by rejecting unfair offers."

Clive Wynne, an associate professor in the psychology department of the University of Florida, isn't so sure the experiment measures the animals reaction to fairness.

"What it means is individuals are responding negatively to being treated less well," he said in a telephone interview.

But the researchers didn't do a control test that had been done in monkey studies, Wynne said, in which a preferred reward was visible but not given to anyone. In that case the monkeys went on strike because they could see the better reward but got something lesser.

Range responded, however, that her team did indeed do that control test as well as others in which food was moved or held in the hand but not given to the dog being tested.

In dogs, Wynne noted, the quality of reward didn't seem to matter, so the test only worked when they got no reward at all.

However, Wynne added, there is "no doubt in my mind that dogs are very, very sensitive to what people are doing and are very smart."

{alt}