Skip to content

Latest commit

 

History

History
19 lines (11 loc) · 877 Bytes

README.md

File metadata and controls

19 lines (11 loc) · 877 Bytes

This is the colorectal adenoma dataset for "CAMEL: A Weakly Supervised Learning Framework for Histopathology Image Segmentation" (ICCV 2019). Paper

Overview

We open-source the colorectal adenoma dataset with image-level labels for weakly supervised learning. This dataset contains 177 whole slide images (156 contain adenoma) gathered and labeled by pathologists from the Department of Pathology, The Chinese PLA General Hospital.

Download

The dataset is hosted on Google Drive or Baidu Drive with password x2o5.

Description

The images are cropped from the whole slide images with size 1280x1280. The image-level label is given as

image_name,label

where the label can be either 1 (adenoma) or 0.