改进扩散桥模型，提升图像生成质量与效率。
Exploring the Design Space of Diffusion Bridge Models
- 通过预处理、端点条件和优化采样算法扩展模型设计空间。
- 在多种图像转换任务中达到最高图像质量和最快采样速度。
- 解决固定条件下的样本多样性不足问题，可定量评估并优化。
扩散桥模型和随机插值方法通过在像素空间中构建分布间的路径，实现高质量的图像到图像（I2I）转换。然而，基于不兼容数学假设的技术大量涌现，阻碍了进展。本文通过为随机插值（SIs）引入预处理、端点条件和优化采样算法，统一并扩展了桥模型的设计空间。这些改进使扩散桥模型在多样I2I任务中同时实现了最先进的图像质量与采样效率。此外，我们识别并解决了此前被忽视的固定条件下样本多样性低的问题，提出了一种输出多样性的量化分析方法，并展示了如何通过调整基础分布进一步提升性能。
原文摘要 · Abstract (English)
Diffusion bridge models and stochastic interpolants enable high-quality image-to-image (I2I) translation by creating paths between distributions in pixel space. However, the proliferation of techniques based on incompatible mathematical assumptions have impeded progress. In this work, we unify and expand the space of bridge models by extending Stochastic Interpolants (SIs) with preconditioning, endpoint conditioning, and an optimized sampling algorithm. These enhancements expand the design space of diffusion bridge models, leading to state-of-the-art performance in both image quality and sampling efficiency across diverse I2I tasks. Furthermore, we identify and address a previously overlooked issue of low sample diversity under fixed conditions. We introduce a quantitative analysis for output diversity and demonstrate how we can modify the base distribution for further improvements.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。