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 2609.34824cs.CV

生成式残差分解

Generative Residual Factorization

Letian Gong, Yuzhou Hong

首次发表
浏览论文内容

中文总结 AI 辅助

本文提出生成式残差分解,将图像块条件分布分解为共享因子后验与残差核,并证明条件熵分解及嵌入预测的不可识别性。

中文摘要 AI 辅助

在共享因子模型下,下一图像块的条件分布可分解为关于共享场景因子的后验分布和给定该因子的残差核。过去信息的充分统计量在后验中替代原始过去信息,但不替代核。条件熵分解为残差熵(任何对因子的观测都无法消除)和后验项(对过去更好的表示可以消除)。下一嵌入预测是浅层映射上的方向似然,因此该映射的纤维不可识别,常数嵌入仍是最小化器。同样的分解在标量高斯模型中是一个等式,并以闭式形式评估。

英文摘要

Under a shared-factor model, the conditional law of the next image patch factors into a posterior over the shared scene factor and a residual kernel given that factor. A sufficient statistic of the past replaces the raw past in the posterior and does not replace the kernel. The conditional entropy splits into residual entropy, which no observation of the factor can remove, and a posterior term, which a better representation of the past can remove. Next-embedding prediction is a directional likelihood on a shallow map, so the fiber of that map is unidentified and a constant embedding remains a minimizer. The same split is an equality in a scalar Gaussian model, evaluated in closed form.

发表机构

  • Zhejiang University of Science and Technology(浙江科技大学)
  • Zhejiang Sci-Tech University(浙江理工大学)

机构由 AI 辅助整理,请以论文原文为准。

补充信息

↑