Datasets

PMC-CancerBox

A multi-modal breast cancer and brain tumour segmentation dataset, released with a representation audit.

DeepFakeX

Eight hundred synthetically generated videos across four manipulation categories, in English and Urdu. The Urdu material is deliberate: almost all existing deepfake datasets are English-only.

The paper

Brain tumour synthetic data

Synthetic brain tumour imaging released for methods work and for augmentation studies.

Protein-drug interaction visualisation

The data files behind the visualisation of protein-drug interactions in lung cancer, written in R.

Vessel segmentation data and code

The datasets and code behind the study of how pre-processing changes retinal vessel segmentation.