Dataset · vector-institute
Medical Images and Captions from PubMed Central
vector-institute/open-pmc-18m
Large-scale English corpus of image-text pairs extracted from open-access PubMed Central medical articles.
What's inside
The dataset contains 10 million to 100 million pairs of images and English text descriptions gathered from PubMed Central scientific publications. Authors from the Vector Institute prepared it for research at the intersection of computer vision and NLP in medicine. The file format in the card is not specified.
What you can do with it
- Pre-training multimodal models (CLIP-like or VLM) in the medical domain
- Training medical image retrieval and caption generation systems
- Languages
- English
- Size
- 10M–100M rows
- Downloads this month
- 44K
- Likes
- 10
How to load it
from datasets import load_dataset
ds = load_dataset("vector-institute/open-pmc-18m")