StyleGAN produces high resolution baby images with parental facial features

babyAC uses an AI technology called StyleGAN to generate images of babies. StyleGAN is an extension of a deep learning model called GAN (Generative Adversarial Network), which allows you to learn the characteristics of two images and combine them to generate anew high-resolution image

babyAC extracts facial features from two parental facial photographs and combines those features with a randomly selected baby photo to create a high-resolution baby face with parental features. Generate a photo

In addition, in order to extract (encode) the features of the parent's face at high speed, we have adopted an encoding technology called pixel2Style2pixel. This reduced the encoding process, which took about a minute, to a few seconds.