期刊
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
卷 44, 期 1, 页码 100-113出版社
IEEE COMPUTER SOC
DOI: 10.1109/TPAMI.2020.3007074
关键词
Edge detection; bi-directional cascade network; scale enhancement; convolutional neural network
资金
- National Key Research and Development Program of China [2018YFE0118400]
- Beijing Natural Science Foundation [JQ18012]
- Natural Science Foundation of China [61936011, 61425025, 61620106009, 61572050, 91538111]
In this paper, a bi-directional cascade network (BDCN) architecture is proposed for edge detection at different scales. The network is supervised at specific scales and utilizes a scale enhancement module (SEM) to generate multi-scale features. The proposed method encourages the learning of multi-scale representations and achieves improved performance in edge detection and other vision tasks.
Exploiting multi-scale representations is critical to improve edge detection for objects at different scales. To extract edges at dramatically different scales, we propose a bi-directional cascade network (BDCN) architecture, where an individual layer is supervised by labeled edges at its specific scale, rather than directly applying the same supervision to different layers. Furthermore, to enrich multi-scale representations learned by each layer of BDCN, we introduce a scale enhancement module (SEM), which utilizes dilated convolution to generate multi-scale features, instead of using deeper CNNs. These new approaches encourage the learning of multi-scale representations in different layers and detect edges that are well delineated by their scales. Learning scale dedicated layers also results in a compact network with a fraction of parameters. We evaluate our method on three datasets, i.e., BSDS500, NYUDv2, and Multicue, and achieve ODS F-measure of 0.832, 2.7 percent higher than current state-of-the-art on the BSDS500 dataset. We also applied our edge detection result to other vision tasks. Experimental results show that, our method further boosts the performance of image segmentation, optical flow estimation, and object proposal generation.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据