---
title: "超越试错：面向图像到视频一致性的智能体优化框架"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.12290v1"
permalink: "https://feed.astralor.com/items/itm_2934e879cfa1627597e84127"
published: "2026-08-12T17:35:16.000Z"
score: 76
channel: "ai"
tags: ["Agentic Self-Improvement","arXiv","Bayesian Optimization","Image-to-Video","Prompt Optimization"]
---

# 超越试错：面向图像到视频一致性的智能体优化框架

## AI 点评

> 这篇论文抓住了视频生成落地中“不可控”的要害，把提示词和参数优化拧成一个闭环，思路挺扎实。对做视频生成产品、或想减少人工调参的团队尤其有用，值得一看。

## 速览

现代黑盒图像到视频（I2V）模型虽然功能强大，但细粒度控制和可靠性不足，随机性导致输出差异大，常需低效试错。本文提出“智能体自改进”框架，将视频合成重构为闭环、目标导向的优化过程。第一阶段通过多模态大模型迭代优化提示词，结合 Davidsonian场景图查询和常见错误问题检测；第二阶段采用贝叶斯优化协同调整随机种子和CFG尺度，并由新的视频-文本一致性评分引导。人工偏好测试中，智能体方法胜率达69%，显著提升视频生成的可预测性和可控性，为可靠的生产级工具铺路。

## 关键点

- 提出智能体自改进框架，将I2V生成视为闭环优化问题，两阶段协同优化提示词与生成参数。
- 创新性引入DSG查询与CMQ检测，并构建VTA评分，实现语义一致性与伪影的自动量化评估。
- 人工偏好测试中胜率达69%，相比未引导搜索显著提升视频质量与一致性。

## 全文

Modern black-box Image-to-Video (I2V) models offer powerful capabilities in automated content creation, yet their lack of fine-grained control and reliability presents significant challenges in professional workflows. Their inherent stochasticity causes minor variations in textual prompts or hyperparameters to yield drastically different outputs often necessitating inefficient, brute-force trial-and-error processes. To address these limitations, we introduce the ``Agentic Self-Improvement" framework, which reframes video synthesis into a closed-loop, goal-directed optimization. Our framework systematically navigates the generation parameter space using a novel two-stage approach. In the first stage, an iterative prompt optimization loop uses a multimodal Large Language Model (mLLM) to refine the input prompt. This refinement implements two automated evaluations: Davidsonian Scene Graph (DSG) queries ensure semantic adherence, and Common Mistake Questions (CMQ) for artifact detection. At the second stage, we use Bayesian optimization to efficiently co-optimize stochastic seeds and CFG scales. This search is guided by a suite of quality metrics, including the novel Video-Text Adherence (VTA) score derived from the DSG and CMQ evaluations. Our framework significantly outperforms unguided search methods: in human preference studies, videos generated via our agentic approach were strongly preferred over baseline outputs, achieving win rates up to 69\%. This work provides a practical and extensible methodology for enhancing the predictability and control of state-of-the-art video generation models, moving the field beyond speculative curiosities toward reliable, production-ready tools.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_2934e879cfa1627597e84127