IOS 15: Identifying People in Photos by Upper Body image

IOS 15: Identifying People in Photos by Upper Body

Apple has recently released a public beta of iOS 15. On Friday, the company announced they used machine learning to better recognize people. The “improved recognition for individuals” was listed as a new feature that will be available in iOS 15. Apple revealed some details of the improved people’s recognition, how it works, and what has changed.

Generally, people in photos are easily identified when the face is clear and visible. But otherwise, identification won’t be possible. According to the Apple company, with iOS 15, the Photos application will be able to effortlessly detect a person even in extreme poses, while they are wearing some accessories, and if their face is occluded.

The company uses a combination of upper body and face to match a particular person when their face isn’t visible. So it doesn’t matter if you stand backward, cover your face, or are in the background, and your image is blurred. You still will be easily detected. This new feature is promised to help a lot with users’ Photos app experience. It is also stressed that despite Apple using machine learning for identifying people, it still does the job of protecting user privacy.

There were also some people concerned about the power and memory consumption this new feature might cause. However, Apple ensured that they had taken steps to minimize it. So this update shouldn’t affect your user experience at all.

According to Apple, faces can be “frequently occluded or simply not visible” due to a variety of reasons. Thus they decided that identifying people by their upper bodies would be possible since “they usually show constant characteristics… These constant characteristics can provide strong clues to identify the person through images captured a few minutes apart.”

Now that the process of people identification has become so much more accessible, it will be easier to search for a certain person in your photos library. We are already excited to test this feature, and what about you?