June 9 (Portaltic/EP) –
google photos ha improved its facial recognition technology increasing the ability to identify people in images, since now also recognizes users who appear from behind and therefore do not show their faces.
The tech giant’s photo storage app features facial recognition technology that analyzes the faces that appear in the photos and tags the people to whom those faces belong. This feature is useful for, for example, automatically creating a folder with all the images in which a specific user appears.
Until now, the ‘app’ has carried out this function through the analysis of the faces of the users. However, the company seems to have improved facial recognition technology powered by Artificial Intelligence (AI) to the point of being able to recognize people with their backs turnedas they have been able to verify from Android Authority.
When a photograph includes a user from behind or in a position where you cannot see your face well, the application does not include people tagged or allow you to add tags manuallysince it does not identify any human being and, therefore, does not consider it necessary.
However, as detailed in this medium, the technology of Google Photos has been able to detect people in positions where they do not show their faces and tag those users correctly in the image. Furthermore, this has occurred both in recently taken images and in old photographs.
In these cases, the tags provided by Google are displayed with the message “face available to add” and also, allows you to change the tagged person or remove the tagin case your identification was not correct.
Despite Google has not shared information about itthis advance in recognition may be due to the fact that the company’s technology is capable of storing information from the entire head of users from videos in which it appears or photos in which the face is shown, even if only part of the face is visible, as Android Authority has pointed out.
Factors such as the similarity of the location with respect to other photographs in which the person’s face does appear, or even other factors such as the environment and the clothing used. All this framed in the Google machine learning capabilities in which it shows continuous progress.
