US 12,494,082 B2
Face image processing method, face image processing model training method, apparatus, device, storage medium, and program product
Keke He, Shenzhen (CN); Junwei Zhu, Shenzhen (CN); Yandan Zhao, Shenzhen (CN); Xu Chen, Shenzhen (CN); Ying Tai, Shenzhen (CN); Chengjie Wang, Shenzhen (CN); Jilin Li, Shenzhen (CN); and Feiyue Huang, Shenzhen (CN)
Assigned to TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED, Shenzhen (CN)
Filed by Tencent Technology (Shenzhen) Company Limited, Shenzhen (CN)
Filed on Nov. 28, 2022, as Appl. No. 18/070,301.
Application 18/070,301 is a continuation of application No. PCT/CN2022/111744, filed on Aug. 11, 2022.
Claims priority of application No. 202110963370.5 (CN), filed on Aug. 20, 2021.
Prior Publication US 2023/0100427 A1, Mar. 30, 2023
Int. Cl. G06V 40/16 (2022.01); G06V 10/80 (2022.01)
CPC G06V 40/168 (2022.01) [G06V 10/806 (2022.01)] 17 Claims
OG exemplary drawing
 
1. A face image processing method performed by a computer device, the method comprising:
acquiring a face image of a source face and a face template image of a template face;
performing three-dimensional face modeling on the face image and the face template image to obtain a three-dimensional face image feature of the face image and a three-dimensional face template image feature of the face template image;
extracting a source face identity feature of the face image from the three-dimensional face image feature;
extracting a template face image feature of the face template image from the three-dimensional face template image feature;
fusing the source face identity feature and the template face image feature to obtain a three-dimensional fusion feature;
performing face replacement feature extraction on the face image based on the face template image to obtain an initial face replacement feature;
transforming the initial face replacement feature based on the three-dimensional fusion feature to obtain a target face replacement feature; and
replacing the template face in the face template image with the source face based on the target face replacement feature to obtain a target face image after the replacement, wherein the target face image preserves identity and shape of the source face and pose and expression of the template face.