Detailing its use of advanced machine learning to significantly improve people recognition in iOS 15, Apple, in a blog post, revealed that the Photos app can now identify people in multiple scenarios, even if their faces are not clear.
To achieve this, Apple matches faces and upper bodies of specific people in the images, reports AppleInsider. The full image is taken as an input, and the detected faces and upper bodies are identified specifically. The faces are then matched with upper bodies to improve individual recognition when facial recognition isn't possible.
"Faces are frequently occluded or simply not visible if the subject is looking away from the camera," Apple wrote.
"To solve these cases, we also consider the upper bodies of the people in the image, since they usually show constant characteristics -- like clothing -- within a specific context. These constant characteristics can provide strong cues to identify the person across images captured a few minutes from each other," it added.
Additionally, the mechanism employs on-device machine learning to ensure privacy. Apple has also taken up several measures to reduce memory and power consumption in the process.
"This latest advancement, available in Photos running in iOS 15, significantly improves person recognition. Using private, on-device machine learning, we can correctly recognise people with extreme poses, accessories, or even occluded faces and use the combination of face and upper body to match the people whose faces are not visible at all," Apple wrote.