Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Friday August 26 2016, @12:11AM   Printer-friendly
from the no-sample-bias dept.

A Baidu voice recognition program has outclassed humans that were typing using smartphone on-screen keyboards:

Computers have already beaten us at chess, Jeopardy and Go, the ancient board game from Asia. And now, in the raging war with machines, human beings have lost yet another battle — over typing. Turns out voice recognition software has improved to the point where it is significantly faster and more accurate at producing text on a mobile device than we are at typing on its keyboard. That's according to a new study by Stanford University, the University of Washington and Baidu, the Chinese Internet giant. The study ran tests in English and Mandarin Chinese.

Baidu chief scientist Andrew Ng says this should not feel like defeat. "Humanity was never designed to communicate by using our fingers to poke at a tiny little keyboard on a mobile phone. Speech has always been a much more natural way for humans to communicate with each other," he says.

Researchers set up a competition, pitting a Baidu program called Deep Speech 2 against 32 humans, ages 19 to 32. The humans took turns saying and then typing short phrases into an iPhone — like "buckle up for safety" and "wear a crown with many jewels" and "this person is a disaster." They found the voice recognition software was three times faster.

Speech Is 3x Faster than Typing for English and Mandarin Text Entry on Mobile Devices (abstract) and full paper (pdf).


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1, Insightful) by Anonymous Coward on Friday August 26 2016, @12:20AM

    by Anonymous Coward on Friday August 26 2016, @12:20AM (#393251)

    ...whilst the notion of having a keyboard has been abandoned.

    Starting Score:    0  points
    Moderation   +1  
       Insightful=1, Total=1
    Extra 'Insightful' Modifier   0  

    Total Score:   1  
  • (Score: 2, Insightful) by Anonymous Coward on Friday August 26 2016, @12:28AM

    by Anonymous Coward on Friday August 26 2016, @12:28AM (#393252)

    First they came for the keyboard, and I did not speak out, because touch screen.

    Then they came for the memory card slot, and I did not speak out, because cloud.

    Then they came for the headphone jack—and no one could speak with me.

    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @03:41AM

      by Anonymous Coward on Friday August 26 2016, @03:41AM (#393322)

      First they came for the keyboard, and I did not speak out, because touch screen.
      Then they came for the memory card slot, and I did not speak out, because cloud.
      Then they came for the headphone jack—and no one could hear me speak.

      FTFY

  • (Score: 3, Funny) by takyon on Friday August 26 2016, @12:30AM

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Friday August 26 2016, @12:30AM (#393254) Journal

    I'm waiting for the follow-up study, Baidu vs. Full QWERTY and Dvorak.

    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @12:45AM

      by Anonymous Coward on Friday August 26 2016, @12:45AM (#393263)
    • (Score: 2) by butthurt on Friday August 26 2016, @01:50AM

      by butthurt (6141) on Friday August 26 2016, @01:50AM (#393287) Journal

      QWERTY and Dvorak weren't designed for entering Chinese characters.

      https://en.wikipedia.org/wiki/Chinese_input_methods_for_computers [wikipedia.org]

      • (Score: 0) by Anonymous Coward on Friday August 26 2016, @02:02AM

        by Anonymous Coward on Friday August 26 2016, @02:02AM (#393291)

        Why can't those barbarians just learn American like the ruling class of the world elite? The gutteral ping-ponging of their local barbarian speech has no place on computers.

      • (Score: 2) by takyon on Friday August 26 2016, @02:08AM

        by takyon (881) <takyonNO@SPAMsoylentnews.org> on Friday August 26 2016, @02:08AM (#393294) Journal

        The study involved English. The study found that voice recognition was 3.0x faster than English hand input, and 2.8x faster than Mandarin Chinese.

        English-speaking users are interested in the performance with respect to the Latin alphabet, not Chinese characters. So my point (or was it a joke?) still stands.

        --
        [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
        • (Score: 2) by butthurt on Friday August 26 2016, @04:07AM

          by butthurt (6141) on Friday August 26 2016, @04:07AM (#393329) Journal

          From the Technology Review article:

          Voice queries are more popular in China because it is more time-consuming to input text, and because some people do not know how to use Pinyin, the phonetic system for transcribing Mandarin using Latin characters.

          I would suppose that this voice recognition system will show its greatest speed and accuracy advantages among people who don't know how to do Pinyin input. Baidu's Mandarin-speaking users (if any) are rather fortunate that the company thought of them.

        • (Score: 0) by Anonymous Coward on Friday August 26 2016, @07:19AM

          by Anonymous Coward on Friday August 26 2016, @07:19AM (#393382)

          Is it faster than entering stuff via stuff like Swiftkey? Since it's only 3X faster than "tapping" on a keyboard I think it's slower.

          It's a matter of whether you want Swiftkey to know your passwords or you prefer some cloud service to know your passwords (lots of this "free" voice recognition stuff have things done at servers).

          I've disabled network access for Swiftkey on my phone (and it works fine- maybe even faster since it doesn't have to send all my stuff to the CIA/NSA, I had problems with Swype lagging a lot presumably sending my data was taking too long :) ). While in theory smartphones could do the voice recognition (and lots of other stuff) themselves, most apps and services prefer to put all that juicy data where they can access it (and presumably make money from it).

  • (Score: 0) by Anonymous Coward on Friday August 26 2016, @12:31AM

    by Anonymous Coward on Friday August 26 2016, @12:31AM (#393255)

    ...and while an expert at Morse code with a fast key beat some texting champions, nearly every one has forgotten that too.

    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @12:38AM

      by Anonymous Coward on Friday August 26 2016, @12:38AM (#393259)

      Obviously. Morse keying requires less muscle movement. It's a twich game.

      What everyone conveniently forgot is how shocked the stupid kids were when they lost to old people.

    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @01:58AM

      by Anonymous Coward on Friday August 26 2016, @01:58AM (#393290)

      Were they texting from a full physical keyboard, or from a 12-key pad or a touchscreen?

    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @07:24AM

      by Anonymous Coward on Friday August 26 2016, @07:24AM (#393383)
      How fast is that expert at entering hash tags? ;)

      p.s. there's no code for # yet...
      • (Score: 2) by butthurt on Saturday August 27 2016, @03:27AM

        by butthurt (6141) on Saturday August 27 2016, @03:27AM (#393843) Journal

        in this baidu stanford experiment there was no punctuation nor capital letters so the comparison to morse code is apt

  • (Score: 5, Insightful) by jmoschner on Friday August 26 2016, @01:22AM

    by jmoschner (3296) on Friday August 26 2016, @01:22AM (#393276)

    This isn't so much that voice is so much better, as typing on a phone is a pain and seems to be getting worse. By the time you get used to a particular phone and the picky behavior of its keyboard and have auto-correct more or less tamed, it is time for a new phone and to start the process over again with another poorly designed UI.

    • (Score: 0) by Anonymous Coward on Friday August 26 2016, @01:35AM

      by Anonymous Coward on Friday August 26 2016, @01:35AM (#393280)

      have auto-correct more or less tamed,

      On the contraryy, auto-correct trained me into the patterrn of tapping space-backspace-space to cancel auto-corect, its how I managed to tap so many spellling errors into this posttt.