Robotics

Toddler robot provides insights into early childhood learning

Toddler robot provides insights into early childhood learning
To learn new words, researchers have found that children use an automatic object-association technique that's very similar to how robots learn
To learn new words, researchers have found that children use an automatic object-association technique that's very similar to how robots learn
View 2 Images
To learn new words, researchers have found that children use an automatic object-association technique that's very similar to how robots learn
1/2
To learn new words, researchers have found that children use an automatic object-association technique that's very similar to how robots learn
The iCub robot was put to the test in picking a new object out of a lineup of familiar ones, and it performed just as well as the human children
2/2
The iCub robot was put to the test in picking a new object out of a lineup of familiar ones, and it performed  just as well as the human children

Kids can pick up new words pretty quickly – just ask any parent who's accidentally sworn in front of their little ones. In order to better understand the mechanics of early learning, a new study tested the object association skills of both children and robots, and found that kids use a robotic technique that may not be based on conscious thought.

As a control group, the team began with human children. The kids, two-and-a-half years old, were tasked with picking a certain toy out of a lineup of either three, four, or five objects, with one of the objects being unfamiliar to them. The aim was to get them to learn the name of the new toy through the process of elimination – the "mutual exclusivity" technique, as the study calls it.

If, for example, a child is presented with three toys – one brown, one yellow and one orange, – they may already know that the brown one is called a rabbit and the yellow one is a duck. When asked "can you show me the giraffe?", the majority of the children were able to figure out that the orange toy was the target.

For the next tests, the researchers put a robot called iCub through the same paces. This little guy is an "embodied neural network" – that is, an artificial intelligence system with deep learning capabilities, wrapped in a robotic body with the proportions of a toddler. iCub has been used in language learning studies in the past, and was even lucky enough to be nominated to carry the torch before the London 2012 Olympic Games.

The iCub robot was put to the test in picking a new object out of a lineup of familiar ones, and it performed just as well as the human children
The iCub robot was put to the test in picking a new object out of a lineup of familiar ones, and it performed  just as well as the human children

The robots were pre-trained to recognize 12 objects, and then, like the human children, were asked to pick out certain ones from a group, with the targets being a mix of both known and novel items. Interestingly, the robots learned the new words in exactly the same way, and performed to the same level as the kids.

"This new study shows that mutual exclusivity behavior can be achieved with a very simple 'brain' that just learns associations between words and objects," says Dr Kaite Twomey, one of the study's authors. "In fact, intelligent as iCub seems, it actually can't say to itself 'I know that the brown toy is a rabbit, and I know that that the yellow toy is a duck, so this new toy must be giraffe', because its software is too simple. This suggests that at least some aspects of early learning are based on an astonishingly powerful association-making ability which allows babies and toddlers to rapidly absorb information from the very complicated learning environment."

The research was published in the journal Interaction Studies.

Source: Lancaster University

1 comment
1 comment
Lbrewer42
The statement, ""This new study shows that mutual exclusivity behavior can be achieved with a very simple 'brain' that just learns associations between words and objects," while fascinating, is flawed scientifically. The flaw is in that the machine they developed can make actions which superficially resemble results of s child in the similar situation.
I can show a video of a fire on a TV screen. I can then put a heater in back of the TV to blow heat towards the observer. This does not mean I have now understood how fire actually works b/c have created a sensory condition of the same type.
My attempt at studying fire by making an artificial one also, and certainly, does not tell me anything about how a real fire and the simulated one are possibly connected in any way.
The problem is our systems of technology are becoming so complex, that we tend to forget not to anthropomorphize them at their base levels.
I applaud the ability to make the man-made system. I am disappointed that - at least all that can be gleaned from the article as stated - is that people are being led to believe this complex system of electronics somehow now let's us know how a human brain works.