Multi-scale pseudo labeling for unsupervised deep edge detection
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Standard
Multi-scale pseudo labeling for unsupervised deep edge detection. / Zhou, Changsheng; Yuan, Chao; Wang, Hongxin; Li, Lei; Oehmcke, Stefan; Liu, Junmin; Peng, Jigen.
I: Knowledge-Based Systems, Bind 280, 111057, 2023.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Multi-scale pseudo labeling for unsupervised deep edge detection
AU - Zhou, Changsheng
AU - Yuan, Chao
AU - Wang, Hongxin
AU - Li, Lei
AU - Oehmcke, Stefan
AU - Liu, Junmin
AU - Peng, Jigen
N1 - Publisher Copyright: © 2023 The Author(s)
PY - 2023
Y1 - 2023
N2 - Deep learning currently rules edge detection. However, the impressive progress heavily relies on high-quality manually annotated labels which require a significant amount of labor and time. In this study, we propose a novel unsupervised learning framework for deep edge detection. It adopts a gradient-based method to generate scale-dependent pseudo edge maps, which match with the hierarchical structure of deep networks. It leverages both the representation learning capability of deep learning, and the simplicity of traditional methods. Experiments on three popular data sets show that the proposed method can suppress non-object edges and reduce the gap with its supervised counterpart due to the introduction of information of various scales and smoothing strategy.
AB - Deep learning currently rules edge detection. However, the impressive progress heavily relies on high-quality manually annotated labels which require a significant amount of labor and time. In this study, we propose a novel unsupervised learning framework for deep edge detection. It adopts a gradient-based method to generate scale-dependent pseudo edge maps, which match with the hierarchical structure of deep networks. It leverages both the representation learning capability of deep learning, and the simplicity of traditional methods. Experiments on three popular data sets show that the proposed method can suppress non-object edges and reduce the gap with its supervised counterpart due to the introduction of information of various scales and smoothing strategy.
KW - Edge detection
KW - Multi-scale modeling
KW - Pseudo labeling
KW - Unsupervised learning
UR - http://www.scopus.com/inward/record.url?scp=85174551668&partnerID=8YFLogxK
U2 - 10.1016/j.knosys.2023.111057
DO - 10.1016/j.knosys.2023.111057
M3 - Journal article
AN - SCOPUS:85174551668
VL - 280
JO - Knowledge-Based Systems
JF - Knowledge-Based Systems
SN - 0950-7051
M1 - 111057
ER -
ID: 372609481