Towards Visually Explaining Variational Autoencoders | IEEE Conference Publication | IEEE Xplore