arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2309.03047cs.CVcs.AI
收藏

结合预训练视觉Transformer与CIDER进行域外检测

Combining pre-trained Vision Transformers and CIDER for Out Of Domain Detection

Pôle Universitaire Léonard de Vinci(列奥纳多·达·芬奇大学中心);reciTAL(reciTAL公司);URCA CReSTIC(兰斯大学CReSTIC研究中心);University of Tours(图尔大学)

Grégor Jouet, Clément Duhart, Francis Rousseaux, Julio Laborde, Cyril de Runz

AI总结:

本文研究预训练视觉Transformer与CNN在域外检测中的性能,发现Transformer开箱即用表现更优,且结合CIDER精炼方法可进一步提升OOD检测性能,为任务设定更强基线。

摘要:

域外(OOD)检测是工业应用中的一个关键组成部分,因为它有助于识别模型何时遇到训练分布之外的输入。大多数工业流程依赖预训练模型用于下游任务,如CNN或视觉Transformer。本文研究了这些模型在域外检测任务上的性能。我们的实验表明,预训练的Transformer模型在开箱即用的情况下实现了更高的检测性能。此外,我们展示了预训练的ViT和CNN可以与诸如CIDER之类的精炼方法相结合,以进一步提高其OOD检测性能。我们的结果表明,Transformer是OOD检测的一种有前景的方法,并在许多情境下为该任务设定了更强的基线。

英文摘要:

Out-of-domain (OOD) detection is a crucial component in industrial applications as it helps identify when a model encounters inputs that are outside the training distribution. Most industrial pipelines rely on pre-trained models for downstream tasks such as CNN or Vision Transformers. This paper investigates the performance of those models on the task of out-of-domain detection. Our experiments demonstrate that pre-trained transformers models achieve higher detection performance out of the box. Furthermore, we show that pre-trained ViT and CNNs can be combined with refinement methods such as CIDER to improve their OOD detection performance even more. Our results suggest that transformers are a promising approach for OOD detection and set a stronger baseline for this task in many contexts

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。

↑