---
id: 20260717-T0-12
title: "新论文提出自校正耦合马尔可夫跳变过程，实现图像理解与生成同步"
title_en: "New paper proposes self-correcting Markov jump process for joint image understanding and generation"
url: https://ai.daily.yangsir.net/daily/20260717-T0-12
issue_date: 2026-07-17
publish_date: 2026-07-16T04:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2607.13188
---

# 新论文提出自校正耦合马尔可夫跳变过程，实现图像理解与生成同步

arXiv 新论文提出了一种自校正耦合马尔可夫跳变过程（Self-Correcting Coupled Markov Jump Processes），旨在让 AI 系统同时进行图像理解和生成，打破传统“理解”与“生成”分离的范式。受人类教师边讲边画启发，该方法让两种模态相互修正与重塑，模拟人类认知的耦合循环。

## English Version

**New paper proposes self-correcting Markov jump process for joint image understanding and generation**

A new paper on arXiv proposes Self-Correcting Coupled Markov Jump Processes, enabling AI systems to perform image understanding and generation simultaneously, breaking the traditional separation of the two. Inspired by how a teacher talks and draws together, the method allows both modalities to correct and reshape each other.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2607.13188)

**详情页**：https://ai.daily.yangsir.net/daily/20260717-T0-12

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*