arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2008.01180cs.CV
收藏

用自然语言描述纹理

Describing Textures using Natural Language

University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校)

Chenyun Wu, Mikayla Timm, Subhransu Maji

摘要:

自然图像中的纹理可以通过颜色、形状、其中元素的周期性以及可以使用自然语言描述的其他属性来表征。在本文中,我们研究的问题,描述纹理的视觉属性在一个新的数据集包含丰富的纹理描述,并进行了系统的研究,目前的生成和判别模型接地语言的图像在这个数据集上。我们发现,虽然这些模型捕捉纹理的一些属性,他们未能捕捉几个组成的属性,如点的颜色。我们通过生成具有不同描述的合成但逼真的纹理来对现有模型进行批判性分析。我们的数据集还允许我们训练可解释的模型,并生成基于语言的解释,说明深度网络学习了哪些区分特征,以进行细粒度分类,其中纹理起着关键作用。我们提出了几个细粒度域的可视化,并表明在我们的数据集上学习的纹理属性提供了对Caltech-UCSD Birds数据集上专家设计的属性的改进。

英文摘要:

Textures in natural images can be characterized by color, shape, periodicity of elements within them, and other attributes that can be described using natural language. In this paper, we study the problem of describing visual attributes of texture on a novel dataset containing rich descriptions of textures, and conduct a systematic study of current generative and discriminative models for grounding language to images on this dataset. We find that while these models capture some properties of texture, they fail to capture several compositional properties, such as the colors of dots. We provide critical analysis of existing models by generating synthetic but realistic textures with different descriptions. Our dataset also allows us to train interpretable models and generate language-based explanations of what discriminative features are learned by deep networks for fine-grained categorization where texture plays a key role. We present visualizations of several fine-grained domains and show that texture attributes learned on our dataset offer improvements over expert-designed attributes on the Caltech-UCSD Birds dataset.

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。

↑