Twitter: ‘Sorry for photo algorithm that may emphasize white people’

Multiple incidents in France: why is the country so restless?

France was startled by several incidents on Thursday. For example, three people were killed in an attack in a church in Nice and...

Terrorism experts: Incidents in France are linked to political conflict

France was startled by several incidents on Thursday. For example, three people were killed in an attack in a church in Nice and...

Researcher for Iltalehti: More terrorist attacks expected in France

President of France Emmanuel Macron said Thursday after the terrorist attack in Nice that France will not bend over due to the Islamist terrorist...

Europe overwhelmed by the second wave of covid-19

Europe is overwhelmed by the onslaught of the second hello of the pandemic coronavirus, which has forced governments to decree curfews, closures of...

Europe is fighting the coronavirus and doubts are growing: Are swabs really effective in curbing the pandemic?

PCR tests are the most popular for diagnosing the virus. But they leave questions unanswered, says Spanish virologist Estanislao Nistal-Villán.As annoying as it...

Twitter CEO Dantley Davis has his excuses offered for a Twitter photo algorithm that may systematically emphasize white people in images of people with different skin tones.

A preview image is a smaller version of a posted photo that has been cropped at a certain point. If someone clicks on the preview, they see the full photo.

The issue came to light when a Twitter user posted as a test image last week that featured both former US President Barack Obama and white US Senator Mitch McConnell. The Twitter algorithm cut the preview on McConnell’s face, not Obama.

Various adjustments to the image did not lead to different results. Other Twitter users tried the experiment on it with different photos and saw the same thing happen.

Davis and his team are now investigating the algorithm, among other things to rule out the possibility that the problem is not due to the contrast in the images, such as the presence of a beard on a white face. “Whatever the reason, this is a problem,” he writes on Twitter.

Davis would consider simply stopping using these types of cropped preview images.

Algorithms that use face recognition are more likely to get it wrong with black faces, partly because they are not trained enough on black people. According to Davis, that’s not the problem here, because Twitter doesn’t use facial recognition for photo previews.

.

trending

Related Articles