Analyzing the latent space of GAN through local dimension estimation for disentanglement evaluation

Jaewoong Choi, Geonho Hwang, Hyunsoo Cho, Myungjoo Kang

Abstract: Highlights•We propose a scheme to estimate local intrinsic dimensions in GAN latent spaces.•Our local dimension estimate provides an upper bound on local semantic perturbations.•We propose a layer-wise unsupervised disentanglement score, called Distortion.•We analyze the layers of the mapping network in StyleGANs through Distortion metric.