Image processing model fusion training method and device, image processing method and device and storage medium

Abstract

The invention provides an image processing model fusion training method. The method comprises the steps of according to a single to-be-replaced face in an original image set, updating parameters of anencoder of a single image processing model and parameters of a decoder of the single image processing model; and under the condition of keeping the parameters of the encoder and the decoder of the single image processing model unchanged, updating the parameters of the encoder of the fused image processing model and the decoder of the fused image processing model according to different faces to bereplaced and different target faces in the original image set. The invention further provides an image processing method and device fusing the image processing model and a storage medium. According to the method, different images can be processed in batches under the condition that the recognition precision of the fusion image processing model is ensured, and storage resources occupied by the fusion image processing model are compressed.

Description

technical field [0001] The invention relates to image processing technology, in particular to a training method for a fusion image processing model, an image processing method, a device and a storage medium. Background technique [0002] In applications such as movie special effects and Internet social networking, if the style of the face (also called the face to be replaced) in the original image set (such as the picture set or video frame set) is maintained, the faces to be replaced are batched The need to replace the face with another character's face (also known as the target face). To this end, artificial intelligence technology (AI, Artificial Intelligence) provides a solution for training an appropriate image processing model to support the above application. Among them, artificial intelligence is the use of digital computers or digital computer-controlled machines to simulate, extend and expand human intelligence, to perceive the environment, acquire knowledge and u...

AI Technical Summary

Problems solved by technology

[0003]The fusion image processing model is trained based on the specific face to be replaced and the target face. When the different faces to be replaced are updated, a single image is required The parameters in the processing model are fully updated, and the efficiency of training a single image processing model is affected, which in turn affects the efficiency of the image processing model being launched in the production environment to achieve the above applications. At the same time, the image processing model that can realize batch processing of images is in The storage complexity is high during deployment, which takes up a lot of storage resources

Images