Paired Image-to-Image Translation Quality Assessment Using Multi-Method Fusion
This addresses the longstanding challenge of quality assessment for image-to-image translation models, though it appears incremental as it builds on existing metrics.
The paper tackles the problem of evaluating synthesized images in image-to-image translation by proposing a Multi-Method Fusion (MMF) model that predicts similarity to ground truth using Image Quality Assessment metrics, achieving automated model ranking without ground truth data.
How best to evaluate synthesized images has been a longstanding problem in image-to-image translation, and to date remains largely unresolved. This paper proposes a novel approach that combines signals of image quality between paired source and transformation to predict the latter's similarity with a hypothetical ground truth. We trained a Multi-Method Fusion (MMF) model via an ensemble of gradient-boosted regressors using Image Quality Assessment (IQA) metrics to predict Deep Image Structure and Texture Similarity (DISTS), enabling models to be ranked without the need for ground truth data. Analysis revealed the task to be feature-constrained, introducing a trade-off at inference between metric computation time and prediction accuracy. The MMF model we present offers an efficient way to automate the evaluation of synthesized images, and by extension the image-to-image translation models that generated them.