Disentangling the Spatial Structure and Style in Conditional VAE

This paper aims to disentangle the latent space in cVAE into the spatial structure and the style code, which are complementary to each other, with one of them zs being label relevant and the other zu irrelevant. The generator is built by a connected encoder-decoder and a label condition mapping network. Depending on whether the label is related with the spatial structure, the output zs from the condition mapping network is used either as a style code or a spatial structure code. The encoder provides the label irrelevant posterior from which zu is sampled. The decoder employs zs and zu in each layer by adaptive normalization like SPADE or AdaIN. Extensive experiments on two datasets with different types of labels show the effectiveness of our method.