Shulou(Shulou.com)11/24 Report--

CTOnews.com October 17 news, recently, Tencent Technology (Shenzhen) Co., Ltd. applied for "facial forgery model training methods, devices, equipment and storage media" patent announced.

According to the patent specification, with the development of artificial intelligence technology, there are more and more research and application in the field of facial (that is, face) forgery detection. At present, facial forgery detection models based on deep learning mostly rely on pseudo-facial images with tag information for supervised training. however, in practical application scenes, pseudo-facial images with tag information are limited. the accuracy of the facial identification model based on limited pseudo-facial image training is not high.

The application discloses a training method, device, equipment and storage medium of a facial forgery identification model, which relates to the technical field of artificial intelligence. The method includes:

The difference between the pseudo face image and the real face image is less than or equal to the threshold value.

Based on the gradient information corresponding to the pseudo-facial image, the pseudo-facial image and the real facial image are fused to get the fused facial image.

Based on the results of forgery detection corresponding to the fused facial image, the facial forgery detection model is trained.

CTOnews.com learned that the patent summary said that the application embodiment can be applied to artificial intelligence, intelligent transportation, assisted driving and other scenarios, and the application can determine the label data corresponding to the fused facial image while generating the fused facial image, and then realize the self-supervised learning of the model based on the fused facial image and label data, without being limited to the scarcity of samples, so as to improve the accuracy of model identification.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.