We introduce SalGAN, a deep convolutional neural network for visual saliency prediction trained with adversarial examples. The first stage of the network consists of a generator model whose weights ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results