PERSONIFICATION CHARACTER GENERATION METHOD BASED ON GENERATIVE ADVERSARIAL NEURAL NETWORK ALGORITHM

摘要

The present invention relates to a personification character generation method based on a generative adversarial neural network algorithm. According to one embodiment of the present invention, the method comprises: (a) a step of allowing a processor to collect each of human, animal, and character image data and store the same in a database; (b) a step of allowing the processor to process the data by assigning an attention value to a predetermined feature area for each of the image data stored in the database; (c) a step of allowing the processor to learn the processed data by using a generative adversarial neural network algorithm; and (d) a step of allowing the processor to assign the attention value to a predetermined feature area for the input user image data and generate a personification character through data transformation using a generative adversarial neural network algorithm when user image data is input from a user terminal. Accordingly, it is possible to create a personification character reflecting the users appearance and characteristics in detail.;COPYRIGHT KIPO 2020