Scholarlink

<p>Traditional Denoising Diffusion Models (DDMs) with UNet backbones are over-parameterized, compromising their effectiveness on limited hardware and in processing gigapixel images. To address this inefficiency, we introduce two Neural Cellular Automata (NCA)-based DDMs: Diff-NCA and FourierDiff-NCA. Leveraging the efficient local communication of NCA, Diff-NCA drastically reduces parameter counts, effectively generating 512 × 512 pathology slices with just 336k parameters. Extending this approach, FourierDiff-NCA integrates <i>Fourier-based diffusion to facilitate early global communication</i>, essential for handling complex datasets such as CelebA. With only 1.1 m parameters, it achieves a <i>more than two times lower FID score of 49.48 compared to the four times larger UNet, which scores 128.2</i>. This performance disparity underscores the utility of NCA-based methods in enhancing parameter efficiency. FourierDiff-NCA also demonstrates versatility by performing tasks such as super-resolution, out-of-distribution image synthesis, and inpainting without task-specific training.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Parameter-efficient diffusion with neural cellular automata

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年5月6日

查看全文
  • John Kalkhof,
  • Arlene Kühn,
  • Yannik Frisch,
  • Anirban Mukhopadhyay

摘要

Traditional Denoising Diffusion Models (DDMs) with UNet backbones are over-parameterized, compromising their effectiveness on limited hardware and in processing gigapixel images. To address this inefficiency, we introduce two Neural Cellular Automata (NCA)-based DDMs: Diff-NCA and FourierDiff-NCA. Leveraging the efficient local communication of NCA, Diff-NCA drastically reduces parameter counts, effectively generating 512 × 512 pathology slices with just 336k parameters. Extending this approach, FourierDiff-NCA integrates Fourier-based diffusion to facilitate early global communication, essential for handling complex datasets such as CelebA. With only 1.1 m parameters, it achieves a more than two times lower FID score of 49.48 compared to the four times larger UNet, which scores 128.2. This performance disparity underscores the utility of NCA-based methods in enhancing parameter efficiency. FourierDiff-NCA also demonstrates versatility by performing tasks such as super-resolution, out-of-distribution image synthesis, and inpainting without task-specific training.