Costume design method based on multi-condition deep convolution generative adversarial network

Abstract

The invention discloses a costume design method based on a multi-condition deep convolution generative adversarial network, and the method comprises the steps: obtaining a costume image data set, andextracting the features of each costume image in the costume image data set; performing unified classification on the features of the garment images in the garment image data set to obtain different types of garment image data subsets, and labeling a category label for each type of garment image data subset; establishing a multi-layer deep convolutional neural network of a generation network and adiscrimination network; sequentially inputting the plurality of garment image data subsets into a multi-layer deep convolutional neural network for training to obtain a garment network model corresponding to each garment image data subset; and inputting demand information into the clothing network model, extracting the clothing network model corresponding to the clothing image data subset, and generating a clothing image in the generation network of the clothing network model.

Description

technical field [0001] The invention belongs to the technical field of clothing design methods, and relates to a clothing design method based on multi-condition deep convolution generation confrontation network. Background technique [0002] With the rapid development of the Internet clothing market, people's shopping methods have undergone tremendous changes, and consumers' demand for clothing styles and quality has also greatly improved compared with the past. It is not an easy task to choose a clothing that suits your taste, the occasion you want to attend, your personality, or even your mood at the time among all kinds of clothing, and it often takes a lot of time and cost. Therefore, in this society that pursues high quality, it is more and more important to efficiently design clothing styles that meet the needs of consumers. [0003] Traditional clothing design methods mainly rely on 3D human body scanning technology to obtain 3D clothing prototypes through physical m...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a clothing design method based on multi-condition deep convolution generation confrontation network, which solves the problem of low design efficiency in the prior art

Images