Learning from Weak and Noisy Labels for Semantic Segmentation
Learning from Weak and Noisy Labels for Semantic Segmentation
复制标题
从弱和嘈杂的标签中学习进行语义分割
DOI：
10.1109/tpami.2016.2552172
复制
发表时间：
2017-02-01
影响因子：
23.6
通讯作者：
Gao, Xin
中科院分区：
文献类型：
--
作者：
Lu, Zhiwu;Fu, Zhenyong;Gao, Xin
弱监督语义分割（WSSS）方法旨在从弱（图像级）而不是强（像素级）标签中学习分割模型。通过避免繁琐的像素级注释过程，它可以利用来自Flickr等媒体共享网站的用户标记图像的无限供应来进行大规模应用。然而，这些“免费”的标签/标签往往是嘈杂的，很少有现有的作品解决学习与弱和嘈杂的标签的问题。在这项工作中，我们把WSSS问题转化为标签降噪问题。具体来说，在将每个图像分割成一组超像素之后，弱的和潜在的噪声图像级标签被传播到超像素级，从而产生高噪声标签;因此，语义分割的关键是识别和校正超像素噪声标签。为此，一种新<inline-formula><tex-math notation="LaTeX">的$L_1$</tex-math><alternatives><inline-graphic xlink:href="xiang-ieq1-2552172.gif"/></alternatives></inline-formula>-优化的稀疏学习模型制定直接和明确地检测噪声标签。为了解决<inline-formula><tex-math notation="LaTeX">$L_1$</tex-math><alternatives><inline-graphic xlink:href="xiang-ieq2-2552172.gif"/></alternatives></inline-formula>-优化问题，我们进一步开发了一个有效的学习算法，通过引入一个中间标记变量。在三个基准数据集上进行的大量实验表明，我们的方法在无噪声标签的情况下产生了最先进的结果，同时在弱标签也有噪声的情况下显著优于现有方法。
A weakly supervised semantic segmentation (WSSS) method aims to learn a segmentation model from weak (image-level) as opposed to strong (pixel-level) labels. By avoiding the tedious pixel-level annotation process, it can exploit the unlimited supply of user-tagged images from media-sharing sites such as Flickr for large scale applications. However, these ‘free’ tags/labels are often noisy and few existing works address the problem of learning with both weak and noisy labels. In this work, we cast the WSSS problem into a label noise reduction problem. Specifically, after segmenting each image into a set of superpixels, the weak and potentially noisy image-level labels are propagated to the superpixel level resulting in highly noisy labels; the key to semantic segmentation is thus to identify and correct the superpixel noisy labels. To this end, a novel <inline-formula><tex-math notation="LaTeX"> $L_1$</tex-math><alternatives><inline-graphic xlink:href="xiang-ieq1-2552172.gif"/></alternatives></inline-formula> -optimisation based sparse learning model is formulated to directly and explicitly detect noisy labels. To solve the <inline-formula><tex-math notation="LaTeX">$L_1$</tex-math><alternatives> <inline-graphic xlink:href="xiang-ieq2-2552172.gif"/></alternatives></inline-formula>-optimisation problem, we further develop an efficient learning algorithm by introducing an intermediate labelling variable. Extensive experiments on three benchmark datasets show that our method yields state-of-the-art results given noise-free labels, whilst significantly outperforming the existing methods when the weak labels are also noisy.