Friday, September 25, 2015

Diego San: Robot Baby Shows Baby Smiles as Control to Make Moms Smile


Summary: Diego San robot baby shows baby smiles as control to make moms smile, not for shared mom-baby smiles, according to a study in Sept. 23's PLOS ONE.


toddler-like robot, Diego San: David Hanson/Machine Perception Laboratory, usage restrictions: credit required, via EurekAlert!

Why do babies smile so much? Is there more to baby smiles than just smiling? Is there more to smiling sessions between mothers and infants than just a happy game? Is it possible for infants under the age of four months to have behavioral goals in social interactions with caregivers, parents, and other family members?
In a study published Sept. 23, 2015, in peer-reviewed, public-access journal PLOS ONE, a team of computer scientists, developmental psychologists and roboticists at the University of California–San Diego (UC San Diego) have sought to demystify baby smiles by programming a baby-like robot with information gleaned from first studying smiling behaviors in 13 pairs of mothers and their infants aged four to 17 weeks. Researchers incorporated the pioneer work by Massachusetts Institute of Technology (MIT) Mathematics Professor Norbert Wiener (Nov. 26, 1894–March 18, 1964) on Cybernetics and Control Theory in understanding goal-oriented, intentional behaviors in animals and machines.
The study identified four smiling scenarios in mother-infant interactions. Mother smiles but infant does not smile. Infant smiles but mother does not smile. Both mother and infant smile. Neither smiles.
The four smiling scenarios were analyzed as driven by four possible goals. One possible goal maximizes maternal smiling time with infant’s simultaneous smiling. Another possible goal maximizes maternal smiling time with unsmiling infant. A third possibility maximizes maternal non-smiling time with infant smiling. A fourth possibility maximizes maternal non-smiling time with unsmiling infant.
The mother-infant smiling interaction study revealed the disparity between infant and maternal goals. The maternal goal maximizes simultaneous smiling between mother and child. The infant’s goal maximizes mother-only smiling time.
Results of the personal mother-child smiling interactions were then programmed into a sophisticated, child-like robot named Diego San for validation in a contingent human-robot interaction study involving 32 UC San Diego undergraduates. The Computer Expression Recognition Toolbox (CERT) guided Diego San in the robot baby’s perception of smiling/unsmiling facial expressions.
Four smiling configurations in the human-robot interactive study corresponded equivalently to the mother-infant smile study. Both robot and human smile in one configuration. Robot baby smiles but human does not smile in a second configuration. In a third configuration, baby robot does not smile but human does smile. In the fourth configuration, neither smiles.
Diego San displayed four control strategies in interactions with participants. One strategy maximizes adult-only smiles as a corollary of the infant control strategy of maximizing mother-only smiles. Another strategy replays smile times clocked by previous participants. A third strategy matches participant’s smiling/unsmiling state by Diego San as a mirror behavior. A fourth control strategy hybridizes maximizing adult-only smiles with increasing likeliness of mirroring by Diego San of participants’ smiling/unsmiling states.
The results of both studies revealed identical preferences by mothers and undergraduates for simultaneous smiling with infants and Diego San, respectively. Both studies also agreed in the goals of maximizing mother-only and adult-only smiles by infants and Diego San, respectively.
Those inscrutable baby smiles seem to be driven by the goal of maximizing mother-only smiles rather than of sharing in simultaneous smiling of mother and child.
Why? Does one-way smiling make babies feel super important?
If only babies could reveal their thoughts, we could truly see the world through their eyes.

Diego San's expressions ~ smiling and unsmiling: Paul Ruvolo et al., CC BY 4.0, via PLOS ONE

Acknowledgment
My special thanks to talented artists and photographers/concerned organizations who make their fine images available on the internet.

Image credits:
toddler-like robot, Diego San: David Hanson/Machine Perception Laboratory, usage restrictions: credit required, via EurekAlert! @ http://www.eurekalert.org/multimedia/pub/99738.php?from=306867
Diego San's expressions ~ smiling and unsmiling ("Infants Time Their Smiles to Make Their Moms Smile," Figure 3): Paul Ruvolo et al., CC BY 4.0, via PLOS ONE @ http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136492

For further information:
"Babies time their smiles to make their moms smile in return." EurekAlert! > Public Releases. Sept. 23, 2015.
Available @ http://www.eurekalert.org/pub_releases/2015-09/uoc--btt092215.php
Patringenaru, Ioana. "Babies Time Their Smiles to Make Their Moms Smile in Return." UC San Diego News Center > Press Release. Sept. 23, 2015.
Available @ http://ucsdnews.ucsd.edu/pressrelease/babies_time_their_smiles_to_make_their_moms_smile_in_return
Ruvolo, Paul, Daniel Messinger and Javier Movellan. “Infants Time Their Smiles to Make Their Moms Smile.” PLOS ONE, Sept. 23, 2015.
Available @ http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136492


No comments:

Post a Comment

Note: Only a member of this blog may post a comment.