Abstract: This study investigates StyleGAN's potential to generate digital camera images of paper fibers (Kozo, Mitsumata, and Gampi) for augmenting deep convolutional neural network (DCNN) training ...