Datasets:
metadata
license: other
language:
- en
task_categories:
- image-segmentation
tags:
- medical-image-segmentation
- medical-imaging
- vision-language
- wpvlmedseg
pretty_name: W-PVLMedSeg Datasets
size_categories:
- 100K<n<1M
W-PVLMedSeg Datasets
This dataset repository contains the biomedical image segmentation datasets used by W-PVLMedSeg, packaged as one zip archive per dataset. The archive contents preserve the project layout:
<DATASET>/
Prompts_Folder/
Train_Folder/
img/
label/
Val_Folder/
img/
label/
Test_Folder/
img/
label/
Download all archives with:
huggingface-cli download DanRuguo/W-PVLMedSeg-Datasets --repo-type dataset --local-dir data_zips
Then unzip each archive into the project data/ directory so paths such as
data/BUSI/Train_Folder/ are available to the training configs.
Contents
- Total archived size: 11.22 GB
- Manifest with archive checksums:
manifest.json
| Dataset | Modality | Organ | Files | Zip | Source |
|---|---|---|---|---|---|
BKAI |
Endoscopy | Colon | 2005 | 1.36 GB | source |
BRISC |
MRI | Brain | 12007 | 783.04 MB | source |
BTMRI |
MRI | Brain | 10141 | 582.86 MB | source |
BUID |
Ultrasound | Breast | 471 | 57.27 MB | source |
BUSBRA |
Ultrasound | Breast | 3757 | 127.72 MB | source |
BUSI |
Ultrasound | Breast | 1573 | 160.19 MB | source |
BUSUC |
Ultrasound | Breast | 1629 | 46.81 MB | source |
ClinicDB |
Endoscopy | Colon | 1231 | 52.69 MB | source |
ColonDB |
Endoscopy | Colon | 767 | 91.41 MB | source |
Covid19 |
X-ray | Chest | 18525 | 209.87 MB | source |
CVC300 |
Endoscopy | Colon | 139 | 16.51 MB | source |
EUS |
Ultrasound | Pancreas | 76341 | 4.58 GB | source |
ISIC |
Dermatoscopy | Skin | 2569 | 2.65 GB | source |
Kvasir |
Endoscopy | Colon | 2013 | 339.57 MB | source |
UDIAT |
Ultrasound | Breast | 333 | 19.40 MB | source |
UWaterlooSkinCancer |
Dermatoscopy | Skin | 454 | 213.18 MB | source |
Notes
These archives aggregate public research datasets and prompt files used for W-PVLMedSeg experiments. Please review and follow the license and usage terms of each original dataset source before redistribution, clinical use, or commercial use.