Pan-cancer nucleus instance segmentation and classification dataset covering ~200,000 nuclei across 5 clinically important classes (neoplastic, non-neoplastic epithelial, inflammatory, connective, dead) and 19 tissue types — the broadest tissue coverage of any public nucleus segmentation dataset. Organised into three cross-validation folds. Used as the primary training and evaluation benchmark for HoVer-Net, CellViT, and most nucleus classification models. Published by the Tissue Image Analytics (TIA) Centre, University of Warwick (Gamper et al. 2019 + 2020). CC-BY-NC-SA-4.0, available from the TIA Centre website.
This entry has no harvestable repository signal (e.g. a dataset, standard, or hosted service), so the deterministic scorer does not assign a ring.
Every value records the source it was harvested from.
| Field | Value | Source |
|---|---|---|
| HuggingFace downloads (monthly) | not_found | huggingface |
| PyPI downloads (monthly) | not_found | pypi |
Recorded deviations — when a harvested value changed between runs.
| When | Field | Was | Now | Source |
|---|---|---|---|---|
| 2026-11-20T21:45:59.780582 | hf_downloads | — | — | huggingface |
| 2026-11-20T21:45:59.724169 | pypi_downloads | — | — | pypi |