标题： 阐明无条件去噪扩散模型中图像的表示 显示英文标题

标题： Elucidating the representation of images within an unconditional diffusion model denoiser

摘要： 生成扩散模型通过估计分数来学习多样化图像数据集上的概率密度，分数由训练用于去除噪声的神经网络估计。 尽管它们在生成高质量图像方面取得了显著成功，但基础分数网络的内部机制尚未被充分理解。 在这里，我们研究了在ImageNet数据集上训练用于去噪的UNet，以更好地理解其内部表示和分数计算。 我们展示了UNet的中间块将单个图像分解为活动通道的稀疏子集，并且这些通道的空间平均值向量可以提供潜在干净图像的非线性表示。 我们开发了一种从这种表示中进行随机重建的新算法，并证明它可以恢复由目标图像表示定义的一组图像中的样本。 然后，我们研究了这种表示的性质，并证明潜在空间中的欧几里得距离对应于由表示诱导的条件密度之间的距离以及图像空间中的语义相似性。 在表示空间中应用聚类算法会产生共享细节（例如，专业化特征、纹理区域、小物体）和全局结构的图像组，但与对象身份的对齐程度仅部分一致。 因此，我们首次表明，仅通过去噪训练的网络包含了图像的丰富且可访问的稀疏表示。 显示英文摘要

摘要： Generative diffusion models learn probability densities over diverse image datasets by estimating the score with a neural network trained to remove noise. Despite their remarkable success in generating high-quality images, the internal mechanisms of the underlying score networks are not well understood. Here, we examine a UNet trained for denoising on the ImageNet dataset, to better understand its internal representation and computation of the score. We show that the middle block of the UNet decomposes individual images into sparse subsets of active channels, and that the vector of spatial averages of these channels can provide a nonlinear representation of the underlying clean images. We develop a novel algorithm for stochastic reconstruction of images from this representation and demonstrate that it recovers a sample from a set of images defined by a target image representation. We then study the properties of the representation and demonstrate that Euclidean distances in the latent space correspond to distances between conditional densities induced by representations as well as semantic similarities in the image space. Applying a clustering algorithm in the representation space yields groups of images that share both fine details (e.g., specialized features, textured regions, small objects), as well as global structure, but are only partially aligned with object identities. Thus, we show for the first time that a network trained solely on denoising contains a rich and accessible sparse representation of images.