We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.IV

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: Domain Adaptive Cascade R-CNN for MItosis DOmain Generalization (MIDOG) Challenge

Abstract: We present a summary of the domain adaptive cascade R-CNN method for mitosis detection of digital histopathology images. By comprehensive data augmentation and adapting existing popular detection architecture, our proposed method has achieved an F1 score of 0.7500 on the preliminary test set in MItosis DOmain Generalization (MIDOG) Challenge at MICCAI 2021.
Comments: updated for MICCAI2021 MIDOG Challenge
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2109.00965 [eess.IV]
  (or arXiv:2109.00965v2 [eess.IV] for this version)

Submission history

From: Ying Cheng [view email]
[v1] Wed, 1 Sep 2021 10:10:28 GMT (550kb,D)
[v2] Wed, 29 Sep 2021 06:24:34 GMT (2081kb,D)

Link back to: arXiv, form interface, contact.