TreeAgent: 林业中基于专家规则与VLM的多智能体自动偏差标注框架

TreeAgent: A Generalizable Multi-Agent Framework for Automated Bias Labeling in Forestry via Compiled Expert Rules and Vision-Language Models

精选理由

林业标注太费人力?TreeAgent用专家规则加VLM自动标注偏差,比传统监督学习强还省心。

AI 摘要

TreeAgent是一个多智能体框架,通过编译专家决策树与视觉语言模型(VLM)实现自动化偏差标注。其解耦声明式决策(D3)框架允许在不同专家定义的决策结构间零修改泛化。在树高偏差分类测试中,TreeAgent的表现优于监督式机器学习基线,并显著减少了专家标注所需的工作量。

AI 翻译 · 中文

TreeAgent是一个多智能体框架,通过编译专家决策树与视觉语言模型(VLM)实现自动化偏差标注。其解耦声明式决策(D3)框架允许在不同专家定义的决策结构间零修改泛化。在树高偏差分类测试中,TreeAgent的表现优于监督式机器学习基线,并显著减少了专家标注所需的工作量。

arXiv cs.AIHuman-labeled data are widely used as reference annotations in ML, despite known variability across annotators in many expert-driven domains. In addition, expert annotation is slow, inconsistent, and remains a major bott