Translations:Diffusion Models Are Real-Time Game Engines/39/zh: Difference between revisions

Latest revision as of 00:23, 9 September 2024

Information about message (contribute)

This message has no documentation. If you know where or how this message is used, you can help other translators by adding documentation to this message.

Message definition (Diffusion Models Are Real-Time Game Engines)

We use DDIM sampling (Song et al., [https://arxiv.org/html/2408.14837v1#bib.bib34 2022]). We employ Classifier-Free Guidance (Ho & Salimans, [https://arxiv.org/html/2408.14837v1#bib.bib12 2022]) only for the past observations condition <math>o_{< n}</math>. We didn’t find guidance for the past actions condition <math>a_{< n}</math> to improve quality. The weight we use is relatively small (1.5) as larger weights create artifacts which increase due to our auto-regressive sampling.

我们使用DDIM采样（Song等人，2022）。我们仅对过去观测条件 $o_{<n}$ 采用了无分类器指导（Ho & Salimans，2022）。我们发现对过去动作条件 $a_{<n}$ 的指导无法提高质量。我们使用的权重相对较小（1.5），因为较大的权重会产生伪影，而我们的自动回归采样则会放大这些伪影。