Translations:Diffusion Models Are Real-Time Game Engines/54/zh: Difference between revisions

Latest revision as of 00:26, 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)

'''Image Quality.''' We measure LPIPS (Zhang et al., [https://arxiv.org/html/2408.14837v1#bib.bib40 2018]) and PSNR using the teacher-forcing setup described in Section [https://arxiv.org/html/2408.14837v1#S2 2], where we sample an initial state and predict a single frame based on a trajectory of ground-truth past observations. When evaluated over a random holdout of 2048 trajectories taken in 5 different levels, our model achieves a PSNR of <math>29.43</math> and an LPIPS of <math>0.249</math>. The PSNR value is similar to lossy JPEG compression with quality settings of 20-30 (Petric & Milinkovic, [https://arxiv.org/html/2408.14837v1#bib.bib22 2018]). Figure [https://arxiv.org/html/2408.14837v1#S5.F5 5] shows examples of model predictions and the corresponding ground truth samples.

图像质量。 我们使用第2节中描述的教师强迫设置来测量LPIPS（Zhang 等人，2018）和PSNR。在该设置中，我们对初始状态进行采样，并根据地面实况的过去观察轨迹预测单帧。在对5个不同级别的2048条随机轨迹进行评估时，我们的模型实现了 $29.43$ 的PSNR值和 $0.249$ 的LPIPS值。PSNR值与质量设置为20-30的有损JPEG压缩相似（Petric & Milinkovic，2018）。图5展示了模型预测和相应地面实况样本的示例。