冻结像素扩散模型可通过自身样本实现自我引导

AI 摘要

该研究提出了一种名为合成自引导（SSG）的方法，使冻结的预训练像素空间扩散模型能够利用自身生成的样本进行自我引导，无需从头训练新模型。通过在中间层附加轻量级预测头，并利用中间层与最终层预测之间的差异作为采样时的引导方向，该方法在ImageNet上显著提升了生成质量，将FID降低超过50%，且适配器训练计算量不到全模型训练的1%。

正文节选

A Frozen Pixel-Space Diffusion Model Can Guide Itself with Its Own Samples Abstract Pixel-space diffusion models aim to learn an end-to-end generator directly over raw pixels. This is challenging because a single model must capture both global structure and local texture in the same high-dimensional space. While recent work improves pixel diffusion through alternative prediction targets, training objectives, and architectures, these advances typically require training a new model from scratch. W

发布时间：—

抓取时间：2026-08-04 20:38

来源机构：Hugging Face