A large-scale 3D face synthesis system with sample similarity suppression

Abstract

The invention discloses a large-scale three-dimensional human face synthesis system with sample similarity suppression, including: a three-dimensional scanning device for obtaining a three-dimensional human face model and its point cloud data; a first processing module for processing the obtained three-dimensional The face model and its point cloud data are preprocessed; the second processing module is used to synthesize the three-dimensional face model to obtain a synthetic three-dimensional face model; the third processing module is used to synthesize the three-dimensional face model and the The sampled 3D face model set is subjected to similarity detection to output the synthesized 3D face model meeting the similarity requirement. On the basis of 3D face model synthesis, this system fully considers the data privacy issues required by the 3D face model synthesis technology, and designs a similarity test model to suppress the similarity between the synthetic model and the real face model. degree, so that the output 3D face synthesis results can meet the special requirements of face data privacy.

Description

technical field [0001] The invention relates to the technical field of three-dimensional face reconstruction, in particular to a large-scale three-dimensional face synthesis system with sample similarity suppression. Background technique [0002] At the World Virtual Reality Industry Conference in October 2018, Professor Huang Tiejun of Peking University pointed out that traditional two-dimensional images and videos are not the real visual media for the human visual nervous system to observe the world. It can be seen that the research on 3D models is more in line with the objective existence of human observation objects. Among common 3D model data, 3D face model data has been widely used in many fields such as games, film and television, and virtual human simulation in simulation systems due to the individual identity representation of 3D face models, the special attributes of personal images and interactive portals. For example, the widely known faceID technology for Apple...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the existing three-dimensional human face synthesis model in the prior art, which only focuses on the optimization model and does not take into account the special requirements for the privacy of human face data required by the three-dimensional human face synthesis model technology, and provides A large-scale 3D face synthesis system with sample similarity suppression, which can make the corresponding 3D face model automatic synthesis results more reliable through similarity inspection and suppress the similarity between the synthetic model and the real face model, and meet the needs of face data Privacy Special Request

Images