AutoVQA-G：用于自动视觉问答与接地标注的自我改进代理框架
AutoVQA-G: Self-Improving Agentic Framework for Automated Visual Question Answering and Grounding Annotation
AI总结：
本文提出AutoVQA-G框架，通过迭代优化流程提升视觉问答接地标注的准确性，优于现有多模态LLM，为构建高质量数据促进更稳健的视觉语言模型训练提供新方法。
摘要：
手动标注高质量的视觉问答与接地（VQA-G）数据集对于推动视觉语言模型（VLMs）的发展至关重要，但难以扩展。现有自动化方法常受限于两个关键问题：（1）由于模型幻觉导致的数据一致性差；（2）基于简单启发法的脆弱验证机制。为解决这些限制，我们引入了AutoVQA-G，一种自我改进的代理框架，用于自动化VQA-G标注。AutoVQA-G采用迭代细化循环，其中一致性评估模块使用链式推理（CoT）进行细粒度视觉验证。基于此反馈，一个记忆增强的提示优化代理分析失败样本的批评，逐步优化生成提示。我们的实验表明，AutoVQA-G生成的VQA-G数据集在视觉接地准确性上优于领先的多模态LLM，为创建高质量数据以促进更稳健的VLM训练和评估提供有前景的方法。代码：https://github.com/rohnson1999/AutoVQA-G
英文摘要：
Manual annotation of high-quality visual question answering with grounding (VQA-G) datasets, which pair visual questions with evidential grounding, is crucial for advancing vision-language models (VLMs), but remains unscalable. Existing automated methods are often hindered by two key issues: (1) inconsistent data fidelity due to model hallucinations; (2) brittle verification mechanisms based on simple heuristics. To address these limitations, we introduce AutoVQA-G, a self-improving agentic framework for automated VQA-G annotation. AutoVQA-G employs an iterative refinement loop where a Consistency Evaluation module uses Chain-of-Thought (CoT) reasoning for fine-grained visual verification. Based on this feedback, a memory-augmented Prompt Optimization agent analyzes critiques from failed samples to progressively refine generation prompts. Our experiments show that AutoVQA-G generates VQA-G datasets with superior visual grounding accuracy compared to leading multimodal LLMs, offering a promising approach for creating high-fidelity data to facilitate more robust VLM training and evaluation. Code: https://github.com/rohnson1999/AutoVQA-G
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。
