Enhancing Diffusion-Based Image Synthesis with Robust Classifier Guidance

Understand

Denoising diffusion probabilistic models (DDPMs) are a recent family of generative models that achieve state-of-the-art results.

• In order to obtain class-conditional generation, it was suggested to guide the diffusion process by gradients from a time-dependent classifier.
• While the idea is theoretically sound, deep learning-based classifiers are infamously susceptible to gradient-based adversarial attacks.
• Therefore, while traditional classifiers may achieve good accuracy scores, their gradients are possibly unreliable and might hinder the improvement of the generation results.