arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.01709cs.CV
收藏

图扩散模型中的偏差缓解

Bias mitigation in graph diffusion models

School of Information Science & Engineering, Lanzhou University(兰州大学信息科学与工程学院)

Meng Yu, Kun Zhan

AI总结:

本文提出方法缓解图扩散模型中的反向起始偏差和暴露偏差,通过改进的Levin采样算法和新的得分修正机制,无需修改网络即可在多个模型和任务中取得最佳效果。

摘要:

现有图扩散模型存在显著的偏见问题。我们发现大多数模型的正向扩散最大扰动分布偏离标准高斯分布,而反向采样始终从标准高斯分布开始,导致反向起始偏见。结合扩散模型固有的暴露偏见,这会降低生成质量。本文提出一种综合方法来缓解这两种偏见。为缓解反向起始偏见,我们采用新设计的Levin采样算法对齐正向最大扰动分布,建立新的反向起始点。为解决暴露偏见,我们引入基于新定义的得分差的得分修正机制。我们的方法无需修改网络,在多个模型、数据集和任务上得到验证,实现了最先进的结果。代码在https://github.com/kunzhan/spp

英文摘要:

Most existing graph diffusion models have significant bias problems. We observe that the forward diffusion's maximum perturbation distribution in most models deviates from the standard Gaussian distribution, while reverse sampling consistently starts from a standard Gaussian distribution, which results in a reverse-starting bias. Together with the inherent exposure bias of diffusion models, this results in degraded generation quality. This paper proposes a comprehensive approach to mitigate both biases. To mitigate reverse-starting bias, we employ a newly designed Langevin sampling algorithm to align with the forward maximum perturbation distribution, establishing a new reverse-starting point. To address the exposure bias, we introduce a score correction mechanism based on a newly defined score difference. Our approach, which requires no network modifications, is validated across multiple models, datasets, and tasks, achieving state-of-the-art results.Code is at https://github.com/kunzhan/spp

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑