HALA, the Roboceptionist
BY MOHAMMAD SHOEB
DOHA: The next time you walk up to the reception of Carnegie Mellon University in Qatar (CMU-Q) and are greeted by the chime of a female robot voice, don’t be surprised. It’s HALA, Qatar’s first robot receptionist. The Roboceptionist, which talks in English and Arabic, will soon be able to identify Arabic-specific accents (similar to language) and read facial expressions, a faculty member associated with the project said recently.
The Roboceptionist will become portable and get more regional looks as there are plans for ‘Hala-2’, which will eventually be taken to the community by the end of this year.
“We are in the process of designing a new body that will make her more portable. If and when we do so, we will be able to take her into the community,” Majd Sakr, Associate Teaching Professor of Computer Science at CMU-Q told The Peninsula in an email interview.
“We have been busy deploying and administering an experiment to measure human perception of politeness versus directness across two groups, native Arabic and native English speakers. Preliminary observations look promising but we’re still analysing the data,” Sakr said.
He added that one of the student projects has focused on identifying Arabic-specific accents (similar to language) in facial expressions.
“What we have found so far is that expressions for positive emotions are easy to identify, however, negative emotions (sadness, anger, disappointment) are harder to recognise. More interestingly, accents in expressions of disappointment seem to be culturally driven,” said Sakr.
HALA, which means ‘hello’ in Arabic, is currently a static roboceptionist placed at the front desk of CMU-Q. It was a joint project of Qatar and CMU’s Pittsburgh Computer Science faculty members and students.
Her interactions are the outcome of many computer programs. Hence, to put HALA in a new setting, she needs to be programmed further to enable her to provide useful answers and have effective interactions.
The Peninsula
