arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.01392cs.CV
收藏

使用标题注释学习视觉表示

Learning Visual Representations with Caption Annotations

NAVER LABS Europe(NAVER LABS 欧洲研究院)

Mert Bulent Sariyildiz, Julien Perez, Diane Larlus

摘要:

预训练通用视觉特征已经成为解决许多计算机视觉任务的关键部分。虽然人们可以在大量注释的ImageNet数据集上学习这些特征,但最近的方法已经研究了允许噪声、更少甚至没有注释来执行这种预训练的方法。从观察到的标题图像很容易抓取,我们认为,这个被忽视的信息源可以被利用来监督视觉表示的训练。为了做到这一点,在语言模型的最新进展的动机下,我们引入{ \em图像条件掩蔽语言建模}(ICMLM)-一个代理任务来学习图像字幕对的视觉表示。ICMLM是依靠视觉线索来预测字幕中的隐藏词。为了解决这个任务,我们提出了混合模型,专用的视觉和文本编码器,我们表明,视觉表示学习作为解决这个任务的副产品转移到各种目标任务。我们的实验证实,图像字幕可以利用注入全球和本地化的语义信息到视觉表示。项目网址:https://europe.naverlabs.com/icmlm。

英文摘要:

Pretraining general-purpose visual features has become a crucial part of tackling many computer vision tasks. While one can learn such features on the extensively-annotated ImageNet dataset, recent approaches have looked at ways to allow for noisy, fewer, or even no annotations to perform such pretraining. Starting from the observation that captioned images are easily crawlable, we argue that this overlooked source of information can be exploited to supervise the training of visual representations. To do so, motivated by the recent progresses in language models, we introduce {\em image-conditioned masked language modeling} (ICMLM) -- a proxy task to learn visual representations over image-caption pairs. ICMLM consists in predicting masked words in captions by relying on visual cues. To tackle this task, we propose hybrid models, with dedicated visual and textual encoders, and we show that the visual representations learned as a by-product of solving this task transfer well to a variety of target tasks. Our experiments confirm that image captions can be leveraged to inject global and localized semantic information into visual representations. Project website: https://europe.naverlabs.com/icmlm.

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。

↑