Skip to content

A few-shot learning benchmark: A dataset of datasets for learning to learn from few examples in the medical imaging domain. Based on the idea of Meta-Dataset.

Notifications You must be signed in to change notification settings

lourencobt/fsMIB

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

fsMIB

A few-shot Medical Imaging Benchmark (fsMIB): A dataset of datasets for learning to learn from few examples in the medical imaging domain. Based on the idea of Meta-Dataset.

User Instructions

Downloading and converting datasets

fsMIB uses datasets that are public and available from different sources and in different formats. Below are the instructions to download and convert the datasets into a common format.

Dataset summary

Dataset (other names) Number of classes Size on disk Converts into
Covid-QU-Ex dataset 3 ~ 3.7 GiB covid [instructions]
Breast Ultrasound Images 3 ~ 0.8 GiB breast [instructions]
Shenzhen set 2 ~ 12 GiB tuberculosis [instructions]
ISIC 2018 7 ~ 14 GiB skinlesion [instructions] \
MURA 2 ~ 11 GiB elbow / finger / forearm / hand / humerus / shoulder / wrist [instructions]

About

A few-shot learning benchmark: A dataset of datasets for learning to learn from few examples in the medical imaging domain. Based on the idea of Meta-Dataset.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages