Written by: Agnetha de Sa, Peak Associate
Anonymizing faces without losing emotion
A team of researchers, which includes those from SFU, are working to improve the way individuals are anonymized for television.
Anonymizing faces in news broadcasts or documentaries usually involves pixelating, blurring, or blacking out an individual’s face. While this common practice leaves the voice intact, it removes a key method of communication — the human face.
Steve DiPaola and Kate Hennessy from SFU’s School of Interactive Arts and Technology (SIAT), along with Taylor Owen, from UBC’s journalism school, are working on a way to use the power of AI “painting” to anonymize an individual’s face. With this technology, facial expressions can be captured along with a voice to give emotion behind a story.
As DiPaola explains, “when artists paint a portrait, they try to convey the subject’s outer and inner resemblance [. . .] with our AI, which learns from more than 1,000 years of artistic technique, we have taught the system to lower the outer resemblance and keep as high as possible the subject’s inner resemblance — in other words, what they are conveying and they are feeling.”
With files from The Star Vancouver.
Salmon jump at the chance to get rid of lice
Two SFU researchers have uncovered a possible explanation for leaping behaviour in salmon.
Previously, researchers have proposed many reasons as to why salmon may leap out of water. Some of these include catching food and avoiding obstacles and predators. Emma Atkinson and John Reynolds, two aquatic ecologists at SFU, recently discovered a new one — that salmon may jump to get rid of lice.
In their field experiment, they studied the leaping behaviour in wild juvenile salmon. In considering juvenile salmon, they knew that previously held ideas about leaping behaviour in salmon would not be true for juvenile salmon since, according to Atkinson, their food source is “almost exclusively” comprised of “underwater zooplankton and their tendency is to scatter rather than leap when escaping from predators.”
Subsequently, Atkinson hypothesized that leaping behaviour in juvenile salmon could be attempts to remove sea lice. By comparing lice counts on two groups of fish, those that were and weren’t allowed to leap, Atkinson and her team found that the leaping salmon had lower seal lice counts compared with their non-leaping counterparts.