Apple has been granted a new patent: machine learning to track users' blood flow and create more realistic virtual portraits

CTOnews.com, Nov. 29 (Xinhua)-- according to a list released today by the U.S. Trademark and Patent Office (USPTO), Apple has obtained a patent for a new technology that uses machine learning technology to track the blood flow of users to create more realistic virtual portraits.

The patent, fully known as "Machine Learning-based Blood Flow Tracking", can be used to infer "blood flow texture maps" through machine learning methods to more realistically present the emotions and facial expressions of virtual portraits.

Existing virtual portrait systems usually do not provide the ability to convey subtle facial representations or emotional states, and Apple's technology, if commercially available, can convey the actual emotional state of users.

Blood flow can be determined by extracting the illumination component because it is shifted from the albedo map. The albedo map describes a facial texture with perfect diffuse light and a static version of the body skin. Therefore, the extracted lighting component indicates the offset of a particular expression from the albedo map.

When you speak, express your emotions, make different facial expressions, or perform any other actions that deform your face, your blood moves around your face differently.

CTOnews.com previously reported that Microsoft has applied for similar technology, through a new machine learning model, avatars or photos can be adjusted to the details to make the photos look more natural.

Microsoft will use a convolutional attention network to improve the accuracy of capturing facial expressions and adjust image details such as blood flow or blushing based on physiological signals such as heart rate.

Related readings:

"Microsoft's new patent announcement: creating a 'blushing' realistic avatar through machine learning."