US 12,299,963 B2
Image processing method and apparatus, computer device, storage medium, and computer program product
Keke He, Guangdong (CN); Junwei Zhu, Guangdong (CN); Hui Ni, Guangdong (CN); Yun Cao, Guangdong (CN); Xu Chen, Guangdong (CN); Ying Tai, Guangdong (CN); Chengjie Wang, Guangdong (CN); Jilin Li, Guangdong (CN); and Feiyue Huang, Guangdong (CN)
Assigned to TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED, Shenzhen (CN)
Filed by TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED, Guangdong (CN)
Filed on Oct. 31, 2022, as Appl. No. 18/051,323.
Application 18/051,323 is a continuation of application No. PCT/CN2022/075952, filed on Feb. 11, 2022.
Claims priority of application No. 202110216698.0 (CN), filed on Feb. 26, 2021.
Prior Publication US 2023/0081982 A1, Mar. 16, 2023
Int. Cl. G06V 10/00 (2022.01); G06V 10/75 (2022.01); G06V 10/774 (2022.01); G06V 10/776 (2022.01); G06V 10/82 (2022.01); G06V 40/16 (2022.01)
CPC G06V 10/774 (2022.01) [G06V 10/751 (2022.01); G06V 10/759 (2022.01); G06V 10/776 (2022.01); G06V 10/82 (2022.01); G06V 40/168 (2022.01); G06V 40/172 (2022.01)] 20 Claims
OG exemplary drawing
 
1. An image processing method, performed by a computer device, the image processing method comprising:
obtaining a training source face image and a training template face image;
performing additional image feature extraction on the training source face image to obtain a source additional image feature corresponding to the training source face image;
performing identity feature extraction on the training source face image to obtain a source identity feature corresponding to the training source face image;
inputting the training template face image into an encoder in a to-be-trained face swapping model for encoding to obtain a face attribute feature;
inputting the source additional image feature, the source identity feature, and the face attribute feature into a decoder in the to-be-trained face swapping model for decoding to obtain a decoded face image;
obtaining a comparative face image, the comparative face image comprising at least one of the training source face image and a standard face image corresponding to the decoded face image, and the standard face image and the training source face image being face images of a same object; and
obtaining an additional image difference between the decoded face image and the comparative face image, and adjusting model parameters of the encoder and the decoder based on the additional image difference to obtain a trained face swapping model, so as to perform image processing according to the trained face swapping model.