One-to-one Mapping for Unpaired Image-to-image Translation
This addresses the problem of ambiguous mappings in image translation for computer vision applications, offering an incremental improvement over existing methods.
The paper tackles the lack of unique or one-to-one mapping in unpaired image-to-image translation by proposing a self-inverse network learning approach, which demonstrates clear improvements over CycleGAN across various datasets, including achieving state-of-the-art results on the cityscapes benchmark.
Recently image-to-image translation has attracted significant interests in the literature, starting from the successful use of the generative adversarial network (GAN), to the introduction of cyclic constraint, to extensions to multiple domains. However, in existing approaches, there is no guarantee that the mapping between two image domains is unique or one-to-one. Here we propose a self-inverse network learning approach for unpaired image-to-image translation. Building on top of CycleGAN, we learn a self-inverse function by simply augmenting the training samples by swapping inputs and outputs during training and with separated cycle consistency loss for each mapping direction. The outcome of such learning is a proven one-to-one mapping function. Our extensive experiments on a variety of datasets, including cross-modal medical image synthesis, object transfiguration, and semantic labeling, consistently demonstrate clear improvement over the CycleGAN method both qualitatively and quantitatively. Especially our proposed method reaches the state-of-the-art result on the cityscapes benchmark dataset for the label to photo unpaired directional image translation.