Facebook's new Face Recognition system DeepFace is as accurate as human

Researchers at Facebook has developed a complex artificial intelligence system which is capable of detecting the faces almost as accurate as human does. Facebook is calling it 'DeepFace' which is capable of rendering a 3D model of any given human face and can identify it on the basis of its availability on the Facebook.

The 'DeepFace' face recognition system identifies a human face correctly from the two given photos 97.25% of the time, which is very close to a human's 97.53% accuracy.

DeepFace, which Facebook researchers detailed in a recent publication, is capable of viewing a photo and then create a 3D rendering of it by using an algorithm called 'Deep Learning'. The Deep Learning artificial intelligence technology uses more than 120 million parameters to analyze a given image before rendering its 3D model accurately.

Abstract of the whole study as detailed by the Facebook Researchers,'In modern face recognition, the conventional pipeline consists of four stages: detect => align => represent => classify. We revisit both the alignment step and the representation step by employing explicit 3D face modeling in order to apply a piecewise affine transformation, and derive a face representation from a nine-layer deep neural network. This deep network involves more than 120 million parameters using several locally connected layers without weight sharing, rather than the standard convolutional layers. Thus we trained it on the largest facial dataset to-date, an identity labeled dataset of four million facial images belonging to more than 4,000 identities, where each identity has an average of over a thousand samples. The learned representations coupling the accurate model-based alignment with the large facial database generalize remarkably well to faces in unconstrained environments, even with a simple classifier. Our method reaches an accuracy of 97.25% on the Labeled Faces in the Wild (LFW) dataset, reducing the error of the current state of the art by more than 25%, closely approaching human-level performance.'

'The social and cultural implications of face recognition technologies are far reaching, yet the current performance gap in this domain between machines and the human visual system serves as a buffer from having to deal with these implications,' the researchers wrote, as quoted by CNNMoney.

The technology is still in the development phase and Facebook has no plans to implement it on the website in near future. But later or sooner, Facebook will implement it to strengthen the security and privacy features of the website.