Multi-scale pseudo labeling for unsupervised deep edge detection

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Dokumenter

  • Fulltext

    Forlagets udgivne version, 4,46 MB, PDF-dokument

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.

OriginalsprogEngelsk
Artikelnummer111057
TidsskriftKnowledge-Based Systems
Vol/bind280
Antal sider15
ISSN0950-7051
DOI
StatusUdgivet - 2023

Bibliografisk note

Funding Information:
This work was supported by National Natural Science Foundation of China (Grant Nos. 12201145 , 12031003 , 62276208 ), China Postdoctoral Science Foundation, China (Grant No. 2022M710862 ), Natural Science Foundation of Guangdong Province, China (Grant No. 2022A1515110336 ), International Development Plan for Young Scientists of Guangdong Province, China (Grant No. 06400802 ), and Postdoctoral Startup Foundation of Guangdong Province, China (Grant No. 62104330 ). This work was conducted when Changsheng Zhou was an visiting scholar at UCPH with Christian Igel. The authors would like to thank Ankit Kariryaa, Mathias Perslev, Yijie Zhang, Lin Zhang and Haiyang Li for their insightful comments.

Publisher Copyright:
© 2023 The Author(s)

ID: 372609481