SolidMark: Evaluating Image Memorization in Generative Models

Recent works have shown that diffusion models are able to memorize training images and emit them at generation time. However, the metrics used to evaluate memorization and its mitigation techniques suffer from dataset-dependent biases and struggle to detect whether a given specific image has been memorized or not. This paper begins with a comprehensive exploration of issues surrounding memorization metrics in diffusion models. Then, to mitigate these issues, we introduce font-variant: small-capsSolidMark, a novel evaluation method that provides a per-image memorization score. We then re-evaluate existing memorization mitigation techniques. We also show that font-variant: small-capsSolidMark is capable of evaluating fine-grained pixel-level memorization. Finally, we release a variety of models based on font-variant: small-capsSolidMark to facilitate further research for understanding memorization phenomena in generative models. All of our code is available at https://github.com/NickyDCFP/SolidMark.