Dual Encoder-Decoder Based Generative Adversarial Networks for Disentangled Facial Representation Learning | IEEE Journals & Magazine | IEEE Xplore