大型视觉-语言模型中的幻觉评估与分析
Evaluation and Analysis of Hallucination in Large Vision-Language Models
AI总结：
针对大型视觉-语言模型中的幻觉问题，提出基于LLM的幻觉评估框架HaELM，性能接近ChatGPT的95%，并分析致因、提出缓解建议。
摘要：
大型视觉-语言模型（LVLMs）最近取得了显著的成功。然而，LVLMs仍然受到幻觉问题的困扰，这限制了其在许多场景中的实用性。幻觉指的是LVLMs响应中不存在于视觉输入中的信息，这带来了潜在的重大后果风险。目前关于LVLMs幻觉评估的研究有限。在本文中，我们提出了基于大型语言模型的幻觉评估（HaELM），一个基于LLM的幻觉评估框架。HaELM达到了与ChatGPT约95%相当的性能，并具有额外优势，包括低成本、可复现性、隐私保护和本地部署。利用HaELM，我们评估了当前LVLMs中的幻觉。此外，我们分析了导致LVLMs幻觉的因素，并提供了有助于缓解幻觉问题的建议。我们的训练数据和人标注幻觉数据将很快公开。
英文摘要：
Large Vision-Language Models (LVLMs) have recently achieved remarkable success. However, LVLMs are still plagued by the hallucination problem, which limits the practicality in many scenarios. Hallucination refers to the information of LVLMs' responses that does not exist in the visual input, which poses potential risks of substantial consequences. There has been limited work studying hallucination evaluation in LVLMs. In this paper, we propose Hallucination Evaluation based on Large Language Models (HaELM), an LLM-based hallucination evaluation framework. HaELM achieves an approximate 95% performance comparable to ChatGPT and has additional advantages including low cost, reproducibility, privacy preservation and local deployment. Leveraging the HaELM, we evaluate the hallucination in current LVLMs. Furthermore, we analyze the factors contributing to hallucination in LVLMs and offer helpful suggestions to mitigate the hallucination problem. Our training data and human annotation hallucination data will be made public soon.
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。
