arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.21794cs.CVcs.AIcs.LG
收藏

广义分布外检测及其在视觉语言模型时代的发展:综述

Generalized Out-of-Distribution Detection and Beyond in Vision Language Model Era: A Survey

The University of Tokyo(东京大学);S-Lab, Nanyang Technological University(南洋理工大学S实验室);Duke University(杜克大学);Salesforce AI Research(Salesforce人工智能研究);Nanyang Technological University(南洋理工大学);University of Wisconsin–Madison(威斯康星大学麦迪逊分校);Tokyo University of Science(东京科学大学)

Atsuyuki Miyai, Jingkang Yang, Jingyang Zhang, Yifei Ming, Yueqian Lin, Qing Yu, Go Irie, Shafiq Joty, Yixuan Li, Hai Li, Ziwei Liu, Toshihiko Yamasaki, Kiyoharu Aizawa

AI总结:

本综述提出广义OOD检测v2框架,统一VLM时代相关任务,聚焦OOD检测与AD,并探讨LVLM进展与未来挑战。

摘要:

检测分布外(OOD)样本对于确保机器学习系统的安全性至关重要,并已塑造了OOD检测领域。同时,其他几个问题与OOD检测密切相关,包括异常检测(AD)、新颖性检测(ND)、开放集识别(OSR)和离群点检测(OD)。为统一这些问题,提出了广义OOD检测框架,对上述五个问题进行了分类学划分。然而,视觉语言模型(VLM)如CLIP已显著改变了研究范式,模糊了这些领域之间的界限,再次使研究者感到困惑。在本综述中,我们首先提出广义OOD检测v2,概括了VLM时代这些领域的演变。我们的框架揭示,随着某些领域的停滞与整合,当前的主要挑战集中于OOD检测和AD。接着,我们强调定义、问题设置和基准方面的显著变化,从而对OOD检测及相关任务的方法进行全面综述,以厘清它们与OOD检测的关系。最后,我们探讨新兴的大型视觉语言模型(LVLM)时代(如GPT-4V)的进展,并总结开放挑战与未来方向。相关资源可在https://github.com/AtsuMiyai/Awesome-OOD-VLM获取。

英文摘要:

Detecting out-of-distribution (OOD) samples is crucial for ensuring the safety of machine learning systems and has shaped the field of OOD detection. Meanwhile, several other problems are closely related to OOD detection, including anomaly detection (AD), novelty detection (ND), open set recognition (OSR), and outlier detection (OD). To unify these problems, a generalized OOD detection framework was proposed, taxonomically categorizing these five problems. However, Vision Language Models (VLMs) such as CLIP have significantly changed the paradigm and blurred the boundaries between these fields, again confusing researchers. In this survey, we first present a generalized OOD detection v2, encapsulating the evolution of these fields in the VLM era. Our framework reveals that, with some field inactivity and integration, the demanding challenges have become OOD detection and AD. Then, we highlight the significant shift in the definition, problem settings, and benchmarks; we thus feature a comprehensive review of the methodology for OOD detection and related tasks to clarify their relationship to OOD detection. Finally, we explore the advancements in the emerging Large Vision Language Model (LVLM) era, such as GPT-4V. We conclude with open challenges and future directions. The resource is available at https://github.com/AtsuMiyai/Awesome-OOD-VLM.

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。

↑