Scholarlink

We propose a method to distill a complex multistep diffusion model into a single-step conditional GAN student model, dramatically accelerating inference, while preserving image quality. Our approach interprets diffusion distillation as a paired image-to-image translation task, using noise-to-image pairs of the diffusion model’s ODE trajectory. For efficient regression loss computation, we propose E-LatentLPIPS, a perceptual loss operating directly in diffusion model’s latent space, utilizing an ensemble of augmentations. Furthermore, we adapt a diffusion model to construct a multi-scale discriminator with a text alignment loss to build an effective conditional GAN-based formulation. E-LatentLPIPS converges more efficiently than many existing distillation methods, even accounting for dataset construction costs. We demonstrate that our one-step generator outperforms cutting-edge one-step diffusion distillation models – SDXL-Turbo and SDXL-Lightning – on the COCO benchmark.

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

Distilling Diffusion Models Into Conditional GANs

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Minguk Kang,
  • Richard Zhang,
  • Connelly Barnes,
  • Sylvain Paris,
  • Suha Kwak,
  • Jaesik Park,
  • Eli Shechtman,
  • Jun-Yan Zhu,
  • Taesung Park

摘要

We propose a method to distill a complex multistep diffusion model into a single-step conditional GAN student model, dramatically accelerating inference, while preserving image quality. Our approach interprets diffusion distillation as a paired image-to-image translation task, using noise-to-image pairs of the diffusion model’s ODE trajectory. For efficient regression loss computation, we propose E-LatentLPIPS, a perceptual loss operating directly in diffusion model’s latent space, utilizing an ensemble of augmentations. Furthermore, we adapt a diffusion model to construct a multi-scale discriminator with a text alignment loss to build an effective conditional GAN-based formulation. E-LatentLPIPS converges more efficiently than many existing distillation methods, even accounting for dataset construction costs. We demonstrate that our one-step generator outperforms cutting-edge one-step diffusion distillation models – SDXL-Turbo and SDXL-Lightning – on the COCO benchmark.