跳至主導覽 跳至搜尋 跳過主要內容

Structured Landmark Detection via Topology-Adapting Deep Graph Learning

  • Weijian Li*
  • , Yuhang Lu
  • , Kang Zheng
  • , Haofu Liao
  • , Chihung Lin
  • , Jiebo Luo
  • , Chi Tung Cheng
  • , Jing Xiao
  • , Le Lu
  • , Chang Fu Kuo
  • , Shun Miao
  • *此作品的通信作者
  • PAII Inc.
  • University of Rochester
  • University of South Carolina
  • Chang Gung Memorial Hospital
  • Ping An Technology (Shenzhen) Co. Ltd.

研究成果: 圖書/報告稿件的類型 › 會議稿件 › 同儕審查

74 引文 斯高帕斯(Scopus)

摘要

Image landmark detection aims to automatically identify the locations of predefined fiducial points. Despite recent success in this field, higher-ordered structural modeling to capture implicit or explicit relationships among anatomical landmarks has not been adequately exploited. In this work, we present a new topology-adapting deep graph learning approach for accurate anatomical facial and medical (e.g., hand, pelvis) landmark detection. The proposed method constructs graph signals leveraging both local image features and global shape features. The adaptive graph topology naturally explores and lands on task-specific structures which are learned end-to-end with two Graph Convolutional Networks (GCNs). Extensive experiments are conducted on three public facial image datasets (WFLW, 300W, and COFW-68) as well as three real-world X-ray medical datasets (Cephalometric (public), Hand and Pelvis). Quantitative results comparing with the previous state-of-the-art approaches across all studied datasets indicating the superior performance in both robustness and accuracy. Qualitative visualizations of the learned graph topologies demonstrate a physically plausible connectivity laying behind the landmarks.

原文英語
主出版品物標題Computer Vision – ECCV 2020 - 16th European Conference, 2020, Proceedings
編輯Andrea Vedaldi, Horst Bischof, Thomas Brox, Jan-Michael Frahm
發行者Springer Science and Business Media Deutschland GmbH
頁面266-283
頁數18
ISBN(列印)9783030585440
DOIs
出版品狀態已出版 - 2020
事件16th European Conference on Computer Vision, ECCV 2020 - Glasgow, 英國
持續時間: 23 08 2020 → 28 08 2020

出版品系列

名字Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
卷12354 LNCS
ISSN(列印)0302-9743
ISSN(電子)1611-3349

Conference

Conference16th European Conference on Computer Vision, ECCV 2020
國家/地區英國
城市Glasgow
期間23/08/20 → 28/08/20

文獻附註

Publisher Copyright:
© 2020, Springer Nature Switzerland AG.

指紋

深入研究「Structured Landmark Detection via Topology-Adapting Deep Graph Learning」主題。共同形成了獨特的指紋。

引用此