Few-shot prototype alignment regularization network for document image layout segementation

Yujie Li

doi:10.1016/j.patcog.2021.107882

2021-07-01

PRE

Shen Fumin (Fumin Shen)

陈

陈立江 (Lijiang Chen)

Pengxiang Gao

DOI：10.1016/j.patcog.2021.107882

原文链接

原文求助

摘要

En 中文

Despite the great performance in layout analysis tasks made by semantic segmentation, they usually need a large number of annotated images for training and are difficult to learn a new category which is ab-sent in the training categories. Meta-learning and few-shot segmentation have been developed to solve the above two difficulties. In this paper, we propose a novel method dubbed Few-Shot Prototype Align-ment Regularization Network (FS-PARN). The FS-PARN method is inspired by recent studies in both met-ric learning and few-shot segmentation, which just need a few annotated images to solve the above two difficulties. Our FS-PARN method can make better use of the information of the support set by metric learning and have a better effect on image segmentation. It learns classification prototype within an em-bedding space and then completes pixel classification by matching each pixel on the query image with the learned prototype. In addition to obtaining high-quality prototypes through metric learning meth-ods, our FS-PARN method also introduces prototype alignment regularization between support and query sets to make segmentation better. Notably, our FS-PARN model achieves the mean-IoU score of 28.8% and 31.7% on the practical document image datasets, i.e. PASCAL-5i, DSSE-200, and Layout Analysis Dataset, for 1-shot and 5-shot settings respectively. (c) 2021 Elsevier Ltd. All rights reserved.

Keyword:

Meta-learning

Few-shot learning

Metric learning

Semantic segmentation