British researchers create robot that can learn simple words by conversing with humans (w/ Video)

June 16, 2012 By Jon Bardin
The iCub robot named DeeChee learning basic language with Professor Chrystopher Nehaniv and Dr Joe Saunders

In an attempt to replicate the early experiences of infants, researchers in England have created a robot that can learn simple words in minutes just by having a conversation with a human.

The work, published this week in the journal , offers insight into how babies transition from babbling to speaking their first words.

The three-foot-tall , named DeeChee, was built to produce any syllable in the English language. But it knew no words at the outset of the study, speaking only babble phrases like "een rain rain mahdl kross."

During the experiment, a human volunteer attempted to teach the robot simple words for shapes and colors by using them repeatedly in regular speech.

At first, all DeeChee could comprehend was an unsegmented stream of sounds. But DeeChee had been programmed to break up that stream into individual and to store them in its memory. Once there, the words were ranked according to how often they came up in conversation; words like "red" and "green" were prized.

DeeChee also was designed to recognize words of encouragement, like "good" and "well done," from its human conversation partner. That feedback helped transform the robot's babble into coherent words, sometimes in as little as two minutes.

If repetition of sounds helps infants learn a language, then it's not surprising that our first words are often mainstays like "mama" and "dada." But why don't we start using common and simple words like "and" or "the" at the same time?

The answer, said study leader Catherine Lyon, a computer scientist at the University of Hertfordshire, is that the words that form the of our language - words like "at," "with" and "of" - are spoken in hundreds of different ways, making them difficult for newbies to recognize. On the other hand, more concrete words like "house" or "blue" tend to be spoken in the same way nearly every time.

Because the study relied on the human volunteers speaking naturally, Lyon said it was crucial that the robot resemble a person. DeeChee was programmed to smile when it was ready to pay attention to its teacher and to stop smiling and blink when it needed a break. (Though DeeChee was designed to have a gender-neutral appearance, humans tended to treat it as a boy, according to the study.)

"When we asked people to talk to the robot as a small child, it seemed to come quite naturally to them," she said. "When they talk to a bit of disembodied software, you don't get the same response."

Paul Goldstein, a psychologist at Cornell University who has also used robots to study infant learning but wasn't involved in this study, said the work was highly innovative - and that if the researchers' theory about language acquisition is correct, they can use robots to prove it.

"If we really think we understand how infants learn," he said, "then we should be able to build a robot that can do it."

Explore further: Robot learns language through 'conversation' with people

More information: Lyon C, Nehaniv CL, Saunders J (2012) Interactive Language Learning by Robots: The Transition from Babbling to Word Forms. PLoS ONE 7(6): e38236. doi:10.1371/journal.pone.0038236

Related Stories

Robots learn to create language

May 17, 2011

( -- Communication is a vital part of any task that has to be done by more than one individual. That is why humans in every corner of the world have created their own complex languages that help us share the goal. ...

'Motherese' important for children's language development

May 6, 2011

(Medical Xpress) -- Talking to children has always been fundamental to language development, but new research reveals that the way we talk to children is key to building their ability to understand and create sentences of ...

Recommended for you

Privacy becomes a selling point at tech show

January 7, 2019

Apple is not among the exhibitors at the 2019 Consumer Electronics Show, but that didn't prevent the iPhone maker from sending a message to attendees on a large billboard.

China's Huawei unveils chip for global big data market

January 7, 2019

Huawei Technologies Ltd. showed off a new processor chip for data centers and cloud computing Monday, expanding into new and growing markets despite Western warnings the company might be a security risk.


Adjust slider to filter visible comments by rank

Display comments: newest first

5 / 5 (2) Jun 17, 2012
how do you justify this over a talking box ? this is definitely a con to make you see something that isn't there.
Didn't you read the article?! The researchers found that people responded more "naturally" to the baby model. It helps with the experiment. Obviously, no one is under the impression it's a real baby! Scientists use "cons" all the time in experiments (e.g. placebo).
5 / 5 (1) Jun 17, 2012
why does this thing have a brace to maintain a standing position when it's obviously not a functioning robot.

Making it look like a "functioning robot" (i.e. more sentient) would induce better responses from participants. People expecting a full robot are going to be disappointed, but that wasn't the intention.
5 / 5 (1) Jun 17, 2012
It's an off the shelf robot. It has the required sensors built in, so they don't need to design a 'box' from scratch. The research is in visual learning, so they aren't going to waste time teaching it to tap dance so it matches someones sci-fi expectations of a robot, they simply use the functionality they need to do the experiment. Other features can be added later if needed.
1 / 5 (1) Jun 17, 2012
Once the speed in learning is increased then most people will begin to see the value in the program design. I just wonder who is going to repay the national debts once these artificial intelligence/robot creatures go mainstream?
1 / 5 (3) Jun 18, 2012
slayerwulfe could be more positive & exercise his/her/its imagination & needs to read carefully
DeeChee can move the fingers of it's right hand while the left remains immobile...
In my designs I use available outputs as cues to code machine states during debug/development then remove some of them later to simplify.

Its conceivable & practical to tie the hand movements range, pattern & speed to aspects of the recognition/classification algorithm the robot goes through when listening. There might be some contextual mapping too so it has a more natural feel - after all, dont humans do that to a fair degree when conversing ?

When we listen to something new, dont we go through (complex) gestural response patterns ?

So to build that in for the developers is a natural and useful feedback mechanism with this robot & more appropriate in line of sight instead of the person interacting having to look over at a clinical screen with numerical/symbolic trace points etc !
1 / 5 (2) Jun 27, 2012
i am he and you are only trying to sell something not create something, i stand firm i know when i'm being conned.
You, slayerwulfe would only say something like that if you had never designed any products that had complex design parameters that needed efficient feedback to observe hierarchies in respect of program modules that are triggered, followed without exceptions and produce an outcome which has some method of being gauged.

It actually takes more effort to con than to connect outputs within a suitable range in the contextual feedback selected, if you can't see that then this forum is not for you...

Frankly slayerwulfe your comments suggest directly you are considerably naive when it comes to the development of complex linguistic based systems... :-(

Also Vendicar_Decarian could do with exercising some of his/her/it's lazy neurons (subject to entropy) or learn appropriate skills to progress same.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.