What can we learn about a generated image corrupting its latent representation?
This work addresses reliability issues in medical imaging applications using GANs, though it is incremental as it builds on existing latent representation analysis.
The paper tackles the problem of predicting image quality in GAN-based image-to-image translation by corrupting latent representations with noise and measuring output variations, demonstrating that the method can identify uncertain parts of synthesized images and unreliable samples for tasks like liver segmentation.
Generative adversarial networks (GANs) offer an effective solution to the image-to-image translation problem, thereby allowing for new possibilities in medical imaging. They can translate images from one imaging modality to another at a low cost. For unpaired datasets, they rely mostly on cycle loss. Despite its effectiveness in learning the underlying data distribution, it can lead to a discrepancy between input and output data. The purpose of this work is to investigate the hypothesis that we can predict image quality based on its latent representation in the GANs bottleneck. We achieve this by corrupting the latent representation with noise and generating multiple outputs. The degree of differences between them is interpreted as the strength of the representation: the more robust the latent representation, the fewer changes in the output image the corruption causes. Our results demonstrate that our proposed method has the ability to i) predict uncertain parts of synthesized images, and ii) identify samples that may not be reliable for downstream tasks, e.g., liver segmentation task.