Entropy-driven Sampling and Training Scheme for Conditional Diffusion Generation

TL;DR

This paper introduces entropy-aware methods to improve conditional image generation in diffusion models, addressing gradient vanishing issues and enhancing sample quality on ImageNet.

Contribution

It proposes entropy-aware guidance scaling during sampling and entropy-aware training objectives to prevent overconfidence, improving conditional diffusion generation.

Findings

Achieved 10.89% FID improvement on ImageNet1000.

Enhanced unconditional FID from 12 to 6.78.

Demonstrated effectiveness of entropy-aware methods in diffusion models.

Abstract

Denoising Diffusion Probabilistic Model (DDPM) is able to make flexible conditional image generation from prior noise to real data, by introducing an independent noise-aware classifier to provide conditional gradient guidance at each time step of denoising process. However, due to the ability of classifier to easily discriminate an incompletely generated image only with high-level structure, the gradient, which is a kind of class information guidance, tends to vanish early, leading to the collapse from conditional generation process into the unconditional process. To address this problem, we propose two simple but effective approaches from two perspectives. For sampling procedure, we introduce the entropy of predicted distribution as the measure of guidance vanishing level and propose an entropy-aware scaling method to adaptively recover the conditional semantic guidance. For training stage, we propose the entropy-aware optimization objectives to alleviate the overconfident prediction for noisy data.On ImageNet1000 256x256, with our proposed sampling scheme and trained classifier, the pretrained conditional and unconditional DDPM model can achieve 10.89% (4.59 to 4.09) and 43.5% (12 to 6.78) FID improvement respectively. The code is available at https://github.com/ZGCTroy/ED-DPM.