Skip to content
Hot eventLive

谱扩散能量条件噪声调度与白化方法

1 reports1 sources3 hr ago updated

Get the story

AI overview

2026年10月7日,arXiv(Artificial Intelligence,一手来源)发表关于谱扩散的能量条件噪声调度与白化方法的论文。论文提出能量自适应噪声调度与白化策略,用于变换域扩散模型,将全局谱白化与按变换系数能量及图像相关能量路径联合调制的噪声分配结合。该方法保持高斯过渡与闭式边缘分布,兼容标准DDPM和DDIM,无需修改扩散架构。CIFAR-10实验中,紧凑DCTdiff U-Net变体的Fréchet Inception Distance从142.48降至100.45。

Generated from reports · updated 3 hr ago

Timeline

Follow the coverage from different angles.

Oct 7, 2026
  1. arXiv · Artificial Intelligence
    谱扩散的能量条件噪声调度与白化方法

    论文提出能量自适应噪声调度与白化策略,用于变换域扩散模型,将全局谱白化与按变换系数能量及图像相关能量路径联合调制的噪声分配结合。该方法保持高斯过渡与闭式边缘分布,兼容标准 DDPM 和 DDIM,无需修改扩散架构。CIFAR-10 实验将紧凑 DCTdiff U-Net 变体的 Fréchet Inception Distance 从 142.48 降至 100.45。

Heat trend

Current heat 9·Comparable peak 10(Oct 7)·Comparable change over 24 hours –

02.557.510Oct7Oct7Oct7Oct7

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.