CNET Crave

CNET Australia Podcast

Thanks for the memories  July 26, 2012

Virtual talking head can express human emotions

About The Author

CNET Editor

Michelle Starr is the tiger force at the core of all things. She also writes about cool stuff and apps as CNET Australia's Crave editor. But mostly the tiger force thing.

The University of Cambridge has created a digital talking face that can express a range of emotions, including anger, happiness and fear.

Giving and receiving voice commands and answers is becoming more common, firstly with in-car GPS units, and then particularly with the arrival of Siri and its clones. But one problem, according to the University of Cambridge, is that a disembodied digital voice is just so impersonal.

So Toshiba's Cambridge Research Lab and the University of Cambridge's Department of Engineering set about putting a face to the voice.

What the team ended up with looks like it comes straight out of Red Dwarf — perhaps the ancestor to the AI with the 6000 IQ, Holly — floating in a sea of blackness and all.

Very appropriately, the head — Zoe — is modelled after actor Zoë Lister (surely that's not an accident), who the team spent days recording as she recited more than 7000 lines in a variety of emotions. This was refined to six basic emotions: happy, sad, tender, angry, afraid and neutral.

How it works is, the user types a phrase for Zoe to say. Six sliders allow you to set the emotions; for example, you could combine happiness and anger, setting them to halfway or full strength, depending on what you want her to convey. Then you can slow down or speed up her speech, giving a pretty large array of tone.

When tested with a group of 20 volunteers, they were able to accurately guess the emotion 77 per cent of the time — more than with the real-life Zoë, with whom the success rate was 73 per cent.

The team sees Zoe being used in the future as a personal assistant, but there are other potential applications as well, because the framework for the face is very light — tens of MBs — which means that it can be incorporated into small devices. It could also enable people to upload their own faces and voices into the program; the team envisions these being used as sort of "face messages" rather than text messages.

The team also believes it could be used with autistic and deaf children, teaching them to read emotions and lips, respectively.

Just so long as it doesn't decide it's bored. Then we might be in for some trouble.

Check out Holly Zoe in action below.


Add Your Comment


Be the first to comment on this story!

Post comment as

Sponsored Links

Recently Viewed Products