A Style-aware Discriminator for Controllable Image Translation
This addresses the need for more flexible and plausible image translation for applications like style interpolation and content transplantation, though it is incremental as it builds on existing GAN-based translation frameworks.
The paper tackles the problem of limited control and poor interpolation in image-to-image translation by proposing a style-aware discriminator that learns a controllable style space, resulting in outperforming state-of-the-art methods on multiple datasets.
Current image-to-image translations do not control the output domain beyond the classes used during training, nor do they interpolate between different domains well, leading to implausible results. This limitation largely arises because labels do not consider the semantic distance. To mitigate such problems, we propose a style-aware discriminator that acts as a critic as well as a style encoder to provide conditions. The style-aware discriminator learns a controllable style space using prototype-based self-supervised learning and simultaneously guides the generator. Experiments on multiple datasets verify that the proposed model outperforms current state-of-the-art image-to-image translation methods. In contrast with current methods, the proposed approach supports various applications, including style interpolation, content transplantation, and local image translation.