Guiding a Diffusion Transformer with the Internal Dynamics of Itself

TL;DR

This paper introduces Internal Guidance (IG), a simple strategy that improves diffusion model training and sampling by auxiliary supervision, leading to state-of-the-art image generation quality.

Contribution

The paper proposes Internal Guidance, a novel auxiliary supervision method during training that enhances diffusion models' efficiency and output quality.

Findings

IG significantly improves FID scores on ImageNet benchmarks.

Combining IG with classifier free guidance achieves new state-of-the-art results.

IG enhances training efficiency and sample quality across various diffusion model baselines.

Abstract

The diffusion model presents a powerful ability to capture the entire (conditional) data distribution. However, due to the lack of sufficient training and data to learn to cover low-probability areas, the model will be penalized for failing to generate high-quality images corresponding to these areas. To achieve better generation quality, guidance strategies such as classifier free guidance (CFG) can guide the samples to the high-probability areas during the sampling stage. However, the standard CFG often leads to over-simplified or distorted samples. On the other hand, the alternative line of guiding diffusion model with its bad version is limited by carefully designed degradation strategies, extra training and additional sampling steps. In this paper, we proposed a simple yet effective strategy Internal Guidance (IG), which introduces an auxiliary supervision on the intermediate layer during training process and extrapolates the intermediate and deep layer's outputs to obtain generative results during sampling process. This simple strategy yields significant improvements in both training efficiency and generation quality on various baselines. On ImageNet 256x256, SiT-XL/2+IG achieves FID=5.31 and FID=1.75 at 80 and 800 epochs. More impressively, LightningDiT-XL/1+IG achieves FID=1.34 which achieves a large margin between all of these methods. Combined with CFG, LightningDiT-XL/1+IG achieves the current state-of-the-art FID of 1.19.