Huawei tech can generate dynamic user avatars based on actions

Huawei has recently patented an amazing technology that can generate user avatars dynamically based on the user’s actions. This seems something different than the already available avatar creation feature.

China’s patent office has published a new Huawei patent with application number CN115810072A, which was applied on December 6, 2022. The patent is titled “method, device, equipment, medium and program product for generating avatar”.

According to the patent details, Huawei phones with this technology could generate dynamic avatars based on the user’s action or motion. Thereafter, it could provide a more vivid and personalized avatar to improve the user experience.

In more technical detail, the technology contains methods, devices, equipment, media, and program products for generating avatars related to the computer field.

The provided method includes determining motion data of a virtual skeleton corresponding to the motion of the user based on an image of the user, the virtual skeleton including a plurality of feature points.

The method also includes determining a set of geometric structures corresponding to a set of target feature points among the plurality of feature points based on the motion data, and generating a dynamic corresponding to the user’s motion by adding the set of geometric structures at the set of target feature points.

Currently, this actions based Huawei user avatars generating technology remains on paper but since it’s published today, we can see its use in future released devices.